The Beginnings of The Ottoman Empire Clive Foss 2022

OUP CORRECTED PROOF – FINAL, 15/12/21, SPi

OX F O R D S T U D I E S I N B Y Z A N T I UM

Editorial Board
jaś elsner catherine holmes
james howard-j ohnston elizabeth jeffreys
hugh kennedy marc lauxtermann
paul magdalino henry maguire
cyril mango marlia mango
claudia rapp jean-p ierre sodini
jonathan shepard

OXFORD STUDIES IN BYZANTIUM

Oxford Studies in Byzantium consists of scholarly monographs and editions on
the history, literature, thought, and material culture of the Byzantine world.
Church Architecture of Late Antique Northern Mesopotamia

Elif Keser Kayaalp
Byzantine Religious Law in Medieval Italy
James Morton
Caliphs and Merchants
Cities and Economies of Power in the Near East (700–950)
Fanny Bessard
Social Change in Town and Country in Eleventh-Century Byzantium
James Howard-Johnston
Innovation in Byzantine Medicine
The Writings of John Zacharias Aktouarios (c.1275–c.1330)
Petros Bouras-Vallianatos
Emperors and Usurpers in the Later Roman Empire
Civil War, Panegyric, and the Construction of Legitimacy
Adrastos Omissi
The Universal History of Stepʻanos Tarōnecʻi
Introduction, Translation, and Commentary
Tim Greenwood
The Letters of Psellos
Cultural Networks and Historical Realities
Edited by Michael Jeffreys and Marc D. Lauxtermann
Holy Sites Encircled
The Early Byzantine Concentric Churches of Jerusalem
Vered Shalev-Hurvitz
The Beginnings of the

Ottoman Empire
CLIVE FOSS
1
1
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Clive Foss 2022
The moral rights of the author have been asserted
First Edition published in 2022
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2021942724
ISBN 978–0–19–886543–8
DOI: 10.1093/oso/9780198865438.001.0001
Printed and bound in Great Britain by
Clays Ltd, Elcograf S.p.A.
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
In Memoriam
MARK WHITTOW
Who would have enjoyed discussing all this
Contents
List of Maps ix
List of Illustrations xi
Introduction1
1. The Homeland of the Ottomans 9
2. The View from Byzantium 99
3. Reconciling the Accounts 135
4. Non-Narrative Sources 141
5. The Overlords 157
6. Osman and his Neighbors 163
7. Western Asia Minor in the 1330s 191
8. The Aftermath 223
9. Final Thoughts 231
Bibliography 245
Index 255
List of Maps
1. The Homeland xv
2. The Aegean region xvi
3. The Marmara region xvii
List of Illustrations
1.1 Ertuğrul’s tomb in Söğüt 11

1.2 Söğüt, seen from the southeast 12
1.3 Söğüt, from the west 12
1.4 Byzantine remains at Günyarık 15
1.5 The Domaniç mountains, with some grazing ground [5] 18
1.6 Looking down on the plain of İnegöl from the Domaniç range 19
1.7 The great forest near Domaniç 19
1.8 The site of Bilecik 24
1.9 Bilecik: Orhan’s mosque 25
1.10 Karacahisar barely visible on top of the hill [10] 30
1.11 The town of İnönü 34
1.12 İnönü’s fortified cave 36
1.13 Looking east from Sultanönü 37
1.14 Harmankaya, home of Mihal 47
1.15 The Sangarius valley, west of Inhisar 50
1.16 Bozaniçkaya on the Sangarius 51
1.17 Taraklı, a goal of raids 51
1.18 Göynük, once an emirate 53
1.19 Mudurnu, once a bishopric 54
1.20 Caravansaray of Mihal beg at Gölpazar 56
1.21 Çakır Pınar 59
1.22 Yarhisar 59
1.23 Yarhisar: Orhan’s mosque 60
1.24 The walls of Ulubat/Lopadion 64
1.25 The fort of Iki Kule 66
1.26 Malagina, overlooking the Sangarius 67
1.27 Walls of Malagina/Akhisar 68
1.28 The Sangarius at Geyve 69
xii List of Illustrations
1.29 Justinian’s bridge over the Sangarius 70

1.30 Tower-like structure on the bridge 71
1.31 Karadin/Trikokkia 71
1.32 Monument to Akçakoca in Kandıra 74
1.33 Looking down on Nicaea 77
1.34 Lascarid tower of Nicaea 78
1.35 Karacakaya, a blockading fort 79
1.36 Apollonia 82
2.1 Nicaea: tower of Andronicus II 106
2.2 Menteşe’s island fortress 107
2.3 Aerial view looking north toward Acropolis of
Sardis. © Archaeological Exploration of Sardis/President
and Fellows of Harvard College, reproduced with permission 112
2.4 Chliara: a Byzantine military base, secure in the mountains
overlooking the plain [40] 114
2.5 Tripolis 115
2.6 The walls of Philadelphia. © History and Art
Collection/Alamy Stock Photo 116
2.7 Walls and citadel of Magnesia © History and Art
Collection/Alamy Stock Photo 116
2.8 Niketiaton 123
2.9 Nicomedia: Byzantine walls 124
2.10 The gulf of Nicomedia 125
4.1 Orhan’s new coinage, struck in Bursa in 1327. © Classic
Numismatic Group, LLC, http://www.cngcoins.com, reproduced
with permission 143
4.2 Birgi’s great mosque (1312) 149
7.1 Peçin, Menteşe’s new capital 193
7.2 Birgi, Aydın’s capital [50] 196
7.3 Aydın’s summer resort 197
7.4 The citadel of Ephesus 198
7.5 The Cayster valley, one source of Aydın’s wealth 199
7.6 Silver gigliato of Theologos (Ephesus). Courtesy of the
Princeton University Numismatic Collection, Department
of Rare Books and Special Collections, Firestone Library 200
7.7 Nif/Nymphaeum: Byzantine and later fortifications [55] 203
List of Illustrations xiii
7.8 Nymphaeum: the Lascarid palace 204

7.9 Bergama, a Karesi capital 205
7.10 Achyraous/“Akira,” Karesi’s first capital 206
7.11 Kütahya: walls 209
7.12 Kütahya citadel [60] 209
7.13 Bursa, Orhan’s capital 211
7.14 Kastamonu: city center, with citadel in background 215
List of Maps xv
1. The Homeland

2. The Aegean region
List of Maps xvii
3. The Marmara region

Introduction
One of the great historical problems at the dawn of the modern age is the
emergence of the Ottomans and consequent collapse of the Byzantine
empire. At first sight, it seems astonishing that an insignificant Turkish
group in a remote corner of Bithynia on the borders of Byzantium should
rise so rapidly from obscurity to domination. When Osman, the epony-
mous founder of a mighty future empire, was born, his people were a tribe
still wandering, or perhaps recently settled, in the land which was to give
birth to their state. A century later, his descendants had crossed into Europe,
soon to overwhelm all their enemies, and on the threshold of becoming a
world power. Close investigation does little to resolve the problem. It has
fascinated scholars in modern times almost as much as the Fall of the
Roman Empire, and with no more satisfactory results: many theories of
varying plausibility have been constructed, but the mystery remains.1 This
chapter has not the ambition to lift the veil which surrounds the origins of
the Ottomans, but merely to suggest a way of approaching the problem, and
present some material rarely considered in this context.
The earliest Ottoman history depends on Turkish chronicles written in
the fifteenth and sixteenth centuries, with scattered information derived
from earlier Byzantine and Arab writers. The oldest source is the Greek his-
torian George Pachymeres (1242–c.1310), a contemporary of Osman who
first mentions him in 1302, and concludes his narrative in 1308. He is fol-
lowed by Nicephorus Gregoras (1295–1359) and John VI Cantacuzene
(1292–1383), who were active in the reigns of Orhan and Murat I. The
Arabic accounts of al-Umari and the observant traveler Ibn Battuta describe
a situation in the early years of Orhan, around 1333–1335.2 They are con-
temporary with the oldest epigraphical, numismatic, and archaeological
evidence: the first dated Ottoman inscription is of 1333, the earliest coins
from the beginning of the reign of Orhan, the first buildings or traces of
them from the years following the conquest of Nicaea in 1331. These are the
contemporary sources, adequate enough, perhaps, for the reign of Orhan
(c. 1320–1360), but revealing little of the crucial half-century previous when
The Beginnings of the Ottoman Empire. Clive Foss, Oxford University Press. © Clive Foss 2022.
DOI: 10.1093/oso/9780198865438.003.0001
2 The Beginnings of the Ottoman Empire
the state had its origins and first grew under the leadership of Osman
(1281–c.1320).3 The Arabic sources present a vivid image of their own time,
but don’t look back, while the Greeks treat the Ottomans only after they
came in contact with the Byzantines; they tell, and no doubt knew, nothing
of their earlier history or of events in other parts of their domain.
If these were the only sources, little would be known of the reign of
Osman, the beginning of Ottoman history (I shall refer to earlier events as
“Ottoman prehistory”).4 Yet much is recounted, and in considerable detail,
by later Turkish narratives that have provided the base for reconstructing
this period.5 The most important of these are Aşıkpaşazade, who wrote
around 1485, the anonymous chronicles edited by Giese compiled about the
same time, and the history of Neşri written about 1490 and largely based
upon the other two.6 Together, they provide what appears to be a full
account of the early conquests and expansion of the Ottomans, with a
wealth of specific places where the events are supposed to have happened.
Such material may be of considerable value, if it represents genuine tradi-
tions, or worthless, if all fabricated later to fill an uncomfortable void. The
early history of Rome might provide an ominous parallel, in which a highly
detailed and dramatic account of three centuries of “history” is to be con-
sidered the unreliable wishful thinking of later writers anxious to provide a
suitable beginning for a great power, and at best a quarry from which some
genuine history or traditions may be extracted. If this is the case here, most
of the later narratives will have to be rejected, the bare skeleton provided by
contemporary sources retained, and the mystery deepened.
It is my aim here to test the accounts of the earliest Ottoman history by
seeking out the places which they mention to determine not only whether
they existed but also whether they are appropriate for the period and events
described. The first is easy enough: the existence of most of the sites has not
been doubted, and a large-scale map will reveal them; for sites that have
disappeared or changed their name, older maps, or accounts of travelers
may profitably be consulted. Yet such identifications might have little mean-
ing: events could have been retrospectively situated in places which were
later important, or a schoolmaster’s fancy could have associated present
remains with places or buildings mentioned in the chronicles. Here, too,
the dawn of Rome provides many discouragingly instructive parallels: the
Alban Mount, well-marked on maps, need not have been the home of the
ancestors of Romulus any more than the Tarpeian rock be taken to authen-
ticate the legend of Tarpeia. More complicated, and potentially more
rewarding, is the search for sites which may be considered appropriate to
Introduction 3
the period. To some extent, this may be done by comparing the sources, and
considering places which also appear in contemporary independent writers
(in these cases, Greek and Arabic) as authentic. Yet these are a small pro-
portion of the total; for the rest, it is necessary to consider the material
record of standing or ruined structures.
Monuments of the first sultans, whether mosques, baths, caravansarays,
castles, or anything else which may be attributed to the period have been
carefully studied in an admirable and indispensable work to which there is
surely nothing to add.7 The studies of E. Hakki Ayverdi make it easy to draw
up a list of buildings from the period, but this, too, has its pitfalls, for as the
author clearly shows, it is easy to state but difficult to prove that a structure
owes its origin to Osman or Orhan. Local tradition is as likely to obscure as
to aid identification; it has a natural tendency to attribute venerable build-
ings or ruins to the first sultans (if they are Islamic; otherwise, Nimrod,
Solomon, the Genoese, or the Jews might have been the builders). It is
therefore necessary to use the care of Ayverdi and other students of Ottoman
architecture in dealing with early monuments.
The warriors of Osman conquered many places from the Infidel, in this
case the Byzantine emperor or his subordinates. By definition, therefore,
most sites of early Ottoman history will be those of the last ages of
Byzantium, and as likely to have Byzantine as Ottoman monuments. Since
many seem to have been taken by siege or stratagem, fortifications might be
expected, solidly built monuments more able than most to withstand the
ravages of time and man. There is thus another category of evidence which
may be brought into account; the Byzantine monuments, especially castles,
which may be dated to the latest period of their rule. Although Byzantine
churches are well-known and can be dated with some accuracy, the castles
have been little studied, but comparative studies have produced a typology
that allows many structures of the thirteenth century to be identified.8 In
the area to be considered, Nicaea alone preserves churches or their remains;
for the rest, the fortifications will be of some importance. The present inves-
tigation, therefore, will attempt to integrate the physical and the written
records of earliest Ottoman history and thus to test the accuracy of the
sources on which it has depended.
Such a study is only one step. Once the sites have been identified,
recorded, and put on the historical map, it is possible to consider their sig-
nificance, and with it the role of geography—of the physical environment—
in the early history and conquests of the Ottomans. Was it suitable, for
example, for the development of a centralized state, or the wanderings of
nomadic tribes, or something else? Most writers have paid least lip service
to geography, often with quite misleading results. Did the Ottomans, for
example, settle on a “high and rather barren tableland,” or on the “grass-
lands” of the frontier, or were they in a broken country with considerable
“vertical range between summer and winter pastures?” The inaccuracy of
the first two could be shown by a good relief map; the other definition needs
to be checked on the spot.9 If this study has any merit, it will be because it is
based on autopsy: I have visited the great majority of the sites mentioned in
the chronicles, recorded the remains in them, and noted their relation to
each other and to the local environment.
Visiting and identifying sites may help to accomplish the first aim of test-
ing the sources, but the consideration of historical geography involves
another real or potential pitfall. Although the mountains and rivers which
form its most powerful component will be the same, many aspects of the
environment now visible may be quite different from those which con-
fronted Osman: forests have been cut down, swamps drained, agriculture,
and with it population increased. Plainly, there is no way to reconstruct the
Bithynian scene in the thirteenth century with the resources available. If one
day studies of lake cores and sediments, of micro-fauna and flora, have been
made, it might be possible to speak with some precision.10 For the moment,
it is necessary to be aware of the problem, and to attempt to reach as far into
the past as possible, relying on old as well as new observations. For this,
I have made extensive use of early travelers, a body which progressively
increases from a trickle in the fourteenth and fifteenth centuries to an
almost unmanageable stream in the nineteenth.11 Through their eyes, a
more accurate image of the past environment may be obtained which, if not
that of Osman’s time, at least is far closer to it than the present somewhat
tamed image which the country presents.
Whatever the success of this effort, the travelers are a delight to read and
the country to visit. In the words of John Macdonald Kinneir, Captain in the
service of the honorable East India Company and political agent at the
Durbar of his Highness the Nabob of the Carnatic, who observed this coun-
try in 1818: “Bithynia is now included in the great province of Anatolia and
governed by a pasha of three tails who resides at Nicomedia; it is a romantic
and beautiful country, intersected by lofty mountains and fertile valleys;
rich in fruits and wine and abounding in forests and fine trees.”12 The pasha
of three tails and most of the forests are long since gone, but the country
remains to provide pleasure and knowledge.
Introduction 5
In treating this material, I have used a chronological arrangement, with

occasional digressions, mainly following the narrative of Aşıkpaşazade, the
most detailed of the Turkish chroniclers, and supplementing it with addi-
tional information to be found in other sources. I have attempted to identify
and describe the places mentioned, using my own observations and those of
my predecessors, and to discuss whatever appropriate remains they may
contain. Consideration of some places give rise to problems of a more gen-
eral kind; these will be treated individually as they occur, and correlated at
the end in a general discussion of the geography and its significance. I shall
begin at the beginning and survey the century from 1261, when the
Byzantine empire was restored in Constantinople, when the Mongols con-
solidated their hold on Asia Minor and when the first tribal Turcoman
states were formed, and continue until 1354, when the Ottoman crossing
into Europe introduced a very different period.
Ideally, a study like this should produce a life-and-times of Osman and
Orhan or a coherent narrative of the rise of the Ottomans, but the sources
don’t permit any such comprehensive treatment. Instead, this work will
consist of a series of related chapters aiming to provide the evidence for the
period it treats, with the aim of seeing what the physical environment and
the historical sources can reveal about the first two Ottomans. Each chapter
can be read independently of the others, though this necessarily involves a
certain amount of repetition, where the same material is presented from
different viewpoints. The survey of the Homeland will be followed by the
Byzantine view of their own decline, then by whatever information can be
derived from other kinds of sources—coins, inscriptions, buildings, and
documents. Two chapters will then attempt some synthesis by considering
Osman in the broader context of the contemporary emirates of western
Asia Minor and the dominant Seljuks and Mongols; and a survey of western
Asia Minor in a period that sees a rare abundance of sources, the 1330s. In
each case, the sources will appear as a kind of raw material, not yet capable
of producing a pleasing narrative. The result, I hope, will be to see the earli-
est Ottomans from a different point of view, without trying to impose a
theoretical framework on the reader.
The project, when first conceived, seemed original, yet I noted that an
earlier traveler of superb acuity, Andreas David Mordtmann, suggested that
such an investigation might be of use. On his way from Eskişehir to Söğüt
in 1858, he wrote: “All these and the following places as far as the sea played
a major role in the history of the first Ottoman sultan, Osman I; and thus
deserve a more thorough investigation by an historian; until now, however,

no one has given himself this trouble.”13 The problems which faced
Mordtmann and a host of others have largely disappeared: paved roads pre-
vent the traveler from sinking in mud and speed his journey; modern
accommodation, even in remote places, spares him the inedible food and
swarms of vermin; and present security has long since eliminated the ban-
dits who struck fear into most. Yet, with some notable exceptions, and those
for quite specific or different purposes, scholars who could most profit from
knowing this geography as a potential key for understanding the history
seem not to have ventured far into Bithynia. This work, therefore, may rep-
resent a step toward the study which Mordtmann recommended.
In 1927, Franz Taeschner made a beginning with a one-week excursion
by automobile into the homeland, investigating Yenişehir, Iznik, Eskişehir,
and Kütahya along with some lesser sites and reported on their Ottoman
remains. This pioneering work, which made use of the Ottoman sources,
paid especial attention to the network of roads.14 Otherwise, the works cited
above show little if any trace of autopsy. The architectural historians, nota-
bly E. Hakki Ayverdi, are an exception; so, to some extent is Prof. Inalcik
who visited the region and inspired an excavation at a site important for this
subject, Karacahisar.
In 2003 appeared a monumental work of French scholarship, La Bithynie
au moyen age, edited by Bernard Geyer and Jacques Lefort. Its nineteen col-
laborators covered every aspect of the subject from pollen to fortifications,
with detailed attention to the geography and the physical remains, from the
seventh century until well into the Ottoman period. There would seem to be
little to add to this massive learning. Yet, there is still room for the present
work, for La Bithynie does not cover the Ottoman homeland south of Bursa
and Nicaea and is not primarily directed toward the early Ottomans or the
historical problems considered here.
This work has a long history. It was undertaken out of curiosity in 1983,
in indirect connection with the survey of medieval castles in Anatolia which
I was directing for the British Institute of Archaeology in Ankara. Rudi
Lindner’s works and conversations provided inspiration and encourage-
ment, as did discussions with Prof. Halil Inalcık, who took much interest in
an early stage of the project. Exploration was greatly facilitated by generous
friends who allowed themselves and their cars to be pressed into service in
remote and unknown regions: my special thanks go to Mr. and Mrs. Rab
Shiell, Consul-general and Mrs. Timothy Gee, and to Stephen Mitchell and
David Barchard for their ever perceptive and helpful comments. The first
Introduction 7
stage of the work (which appears here as the long introductory chapter on
the Homeland) was completed in the tranquility of All Souls College,
Oxford, where I spent the academic year 1983/84. It benefited from com-
ments of Mme Irene Beldiceanu and especially from the meticulous atten-
tion of the late professor Victor Ménage whose detailed suggestions and
corrections saved me from many mis-statements and errors. Then the
incomplete project was set aside and not revived for thirty years. I have
brought the references up to date as far as possible, but left the descriptions
of sites as they were in 1983, before the drastic transformations of the coun-
tryside in recent decades.
In 2015, an invitation to address a conference in Nicaea provided the
occasion to return to this project, which insensibly turned from a chapter
into a book where the original material could be put into a broader context.
For that, I owe thanks to Kutlu Akalın who organized the meeting and to my
indefatigable and enormously helpful driver Çağla Altıntaş. I am grateful to
David Mitten and Laura Johnson for help with practical matters and to Julian
Baker and Lutz Ilisch for answering numismatic questions. My thanks to all
these friends and colleagues and especially to the villagers of the Homeland
who willingly shared their knowledge of the antiquities in their midst.
Cambridge MA and Oxford
August 2019
Notes
1. The main comprehensive treatments of the rise of the Ottomans, which present
variety of approaches, are Gibbons 1916, Langer and Blake 1924, Köprülü 1935,
Wittek 1932, Arnakis 1947, Lindner 1983 and 2007, Kafadar 1995, and
Lowry 2003. They will be discussed in the concluding chapter, p. 236f.
2. Al-Umari reproduces detailed information from Haydar al-Uryan of Sivrihisar
and Balban the Genoese; see p. 191f.
3. Most of the dates presented here from the Turkish tradition are arbitrary; PRINT
see p. 90 n. 47.
4. Taeschner 1928, 86, calls the period before the conquest of Bursa “die Urgeschiche
des osmanschen Turkentums.”
5. The sources are discussed in all the works cited by Lowry 2003; for more specific
criticism, see Menage 1964.
6. PRINT: See below, p. 9.
7. I refer to the elegant and comprehensive works of Ayverdi 1966, 1972, and 1974.
See also Kuran 1968 for convenient summaries, and Kiziltan 1958.
8. See Foss and Winfield 1986, 150–9 and, for this region, Grelois 2003, 209–24.
9. The quotes are respectively from Langer and Blake 1932, 494, Pitcher 1972, 36
and Lindner 1983, 30.
10. An important beginning has now been made in the chapters on environmental
history in Geyer-Lefort 2003, 153–205.
11. For the travelers, see Saint-Martin 1846, III.710–808 and for an admirably com-
prehensive study of those through 1600, see Yerasimos 1991.
12. Kinneir 1818, 257; the pasha actually resided at Kütahya.
13. Mordtmann 1925, 549.
14. Taeschner 1928, 86–104.
1
The Homeland of the Ottomans
The Sources
The history of the earliest Ottomans is based on three main sources: a group
of anonymous chronicles, the comprehensive narrative of Aşıkpaşazade,
and the history of Neşri. All are products of the late fifteenth century but
incorporate earlier material. Aşıkpaşazade (henceforth APZ) draws on the
Anonymous as well as the lost work of Yahşı Fakih, the son of Orhan’s
imam, whose text APZ used when he had fallen ill and was staying with the
fakih at Geyve on the Sangarius in 1413. Nothing further is known of Yahşı,
but Aşıkpaşazade was a well-known dervish whose experience involved
military campaigns in Anatolia and Thrace and who reflected the views of
frontier warriors. He apparently wrote his chronicle in the context of the
sultan Bayezid II’s request for a history of the Ottoman dynasty and its
achievements. He finished his chronicle in 1484, but it contains the much
earlier material of Yahşı Faqih whose own work ended in 1389 or 1402.
The historian Neşri, who belonged to a more highly educated milieu,
produced in the late fifteenth century a wide-ranging work that drew on
both the Anonymous and APZ. The story is much more complicated than
this, with numerous other works associated or derived from these three.1
In addition, two poets contribute to the subject. Notable among them is
Ahmedi (c.1335–1412), the greatest poet of his age. His 8000-verse epic,
written in the early fifteenth century, devotes 337 verses to the Ottomans,
and only six of them to Osman.2 Enveri finished his verse chronicle in 1465.
It deals with the career of Umur of Aydin (1309–1348), a contemporary of
Osman who, however, never appears in the story.3 Virtually nothing is
known of these authors’ lives.
The account of APZ will provide the framework for the present discus-
sion since it is comprehensive and available in facsimile text, modern
Turkish, and German.4 The Anonymous and Neşri texts will be drawn upon
for divergent material; the three sources together will normally be presented
as the Tradition.
DOI: 10.1093/oso/9780198865438.003.0002
Ottoman Prehistory (APZ cap. 2)
The record of the tribe of Kayi Turks before the time of the eponymous
Osman is recounted in outline by sources whose narrative may be fable, tra-
dition, or history, or some combination of all three.5,6
It may contain elements of historical fact, but little of it can be tested.
That Suleymanşah, the leader of the tribe, brought his followers from Mahan
in Iran to eastern Anatolia at the time of the Mongol invasion is inherently
plausible and well suits the circumstances of the time. That he drowned in
the Euphrates, where a castle and a Turkish enclave in Syria commemorate
the event, is certainly an ancient tradition. From there, the people moved in
the direction of their future home under the leadership of Ertuğrul. He is
supposed to have received land on the Karacadağ near Ankara from the
Seljuk Sultan Ala ad-Din Kayqubad, and later, as a reward for his assistance
in arms, an area with winter and summer pastures on the far western fron-
tier, in and around Söğüt in Bithynia. He settled here with his people and
died in 1281. Ottoman history was in its infancy and these lands, its cradle,
may be seen.
The figure of Ertuğrul remains shadowy, and may be no more real than
that of Romulus. He appears universally in the Turkish tradition, and rarely
anywhere else.7 Yet the countries he inhabited were substantial, even per-
haps one from “prehistory.” Ala ad-Din Kayqubad I (1220–1237) certainly
existed, the most illustrious of the Seljuks of Rum; so illustrious, in fact, that
any number of buildings or events can easily be attributed to him. There is a
Karaca Dağ not near, but about fifty miles south, of Ankara. It is not espe-
cially high, but large, and stands on the edge of the central Anatolian pla-
teau, the natural home of horseman and nomad. It is a plausible site for an
early tribal home, but there is no external evidence to connect the two.
Another specific item is similarly incapable of verification. Neşri reports
that the young Osman in his father’s lifetime, saw and fell in love with Mal
Hatun, daughter of Sheikh Edebali, at the village of It Burnu on the road
between Söğüt and Eskişehir.8 The place still exists, but in itself provides no
confirmation for the story which could have been set there for reasons
unknown at any time in the two centuries between event and narrative.9
The site of the first homeland, which brings the traditions into history, is
of far greater significance. Ertuğrul is supposed to have received a winter
pasture at Söğüt, and summer pastures in the Ermeni Beli (Armenian Pass)
and the Domaniç mountain. These may be identified and visited, and are
worth considering in some detail.
The Homeland of the Ottomans 11
The First Settlement (APZ cap. 2)
The identity of SÖĞÜT with the town that still bears the name has never
been in doubt, its fame assured by the tomb of Ertuğrul (Fig. 1.1) venerated
there for at least five hundred years. The site, often visited and described, is
especially scenic and characteristic of the region.10 Söğüt is built on sloping
hillsides above a stream; its older parts have little level ground (Figs. 1.2, 1.3).
The town is surrounded by steep hills on all sides except the north where
the wheat fields beyond the tomb of Ertuğrul soon drop off sharply to the
valley of the Sangarius, offering magnificent views of the bare rock moun-
tains beyond. The approaches to the town are all difficult, involving passage
through mountains with woods and narrow steep passes. The place appears
to be in a secure position, remote from all. In fact, though, the site is not at
all insignificant, because major routes ran through this rough country.
The earliest description is of Evliya Çelebi, who in the late seventeenth
century noted that Söğüt had seven hundred tile-roofed houses, many
mosques, hans, and baths, and a market; the tomb of Ertuğrul outside the
town was not very prepossessing; a specialty of grape pickles gave additional
local fame.11 Later travelers add various details about the town, and give an
image of the surrounding country which in the nineteenth century was
Fig. 1.1 Ertuğrul’s tomb in Söğüt

Fig. 1.2 Söğüt, seen from the southeast
Fig. 1.3 Söğüt, from the west
productive of silk, grain, and fruit: to the north was an open undulating
country with wheat in the plains and mulberries on the hillsides, while the
slopes of the Söğüt Dağ on the south were noted for their fine gardens,
orchards, woods, and springs.12 Little has changed. This country is typical of
the “middle” of the three geographical zones into which northwest Anatolia
may be divided. The first stretches from Istanbul to the gorges of the
Sangarius, and offers much flat land, some of it swampy, capable now of
intense cultivation of market gardens, with fruit and vegetables of all
kinds and long stands of poplars; in the past much of it was heavily
forested. The middle zone, the broken country between the plains and the
plateau, has poppies, grapes, and mulberries with cypresses and plane
trees in the low-lying areas, nut trees, lindens, and oaks on the slopes, and
a once dense forest of beech, oak, and fir on the mountains. Beyond it
begins the open steppe, suitable for sowing with wheat or for grazing
sheep and cattle.13
Söğüt is mentioned most often in association with the roads, for, as Haci
Kalfa noted in the mid-seventeenth century, it lay on the great route across
Anatolia. This led from Istanbul through Nicaea, Lefke, and Bilecik to
Eskişehir and the plateau, forming part of the major transverse highway
which connected the capital with Syria and the East.14 It was certainly in use
in the sixteenth century, but became far more important in the seventeenth,
especially after Köprülü Mehmet Pasha built the great caravansaray of
Vezirhan, between Lefke and Bilecik, in 1660. Ozuyuk travelers’ accounts
tell much of this and other roads which radiated from Söğüt.
Travelers of the early nineteenth century, who came from the northwest
and left detailed descriptions, followed a direct route from Vezirhan rather
than the traditional (and modern) road along the Kara Su valley and
through Bilecik. This climbed a ridge, then passed through “wild scenery of
broken rocks and barren downs with little or no wood” or “a barren bleak
tract with deep winding valleys” until a pass which was followed by twelve
or fifteen miles of pleasant country with mulberries and grain.15 Beyond
Söğüt, a chain of rocky hills and a long defile led to a bleak and open coun-
try which gradually dropped down to the plain of Eskişehir.
Less frequented routes led south directly to İnönü and southwest to
Bozüyük. The former began with a barren country of volcanic rocks then
passed most of its course in an extensive mountain forest of oak, fir, and
plane until it reached the broad plain of the Sari su.16 The road to Bozüyük
makes a steep climb with spectacular views back over the mountains beyond
the Sangarius, then enters a high upland plain, the yayla of Günyarık before
descending steeply beside the prehistoric acropolis of Bozüyük. Another
route to the north, passable now but apparently not mentioned by travelers,
winds down long and steep slopes through some large and prosperous
villages to the narrow plain of the Sangarius some 500 meters below. There
is no good route to the west: only tracks lead through a wild and broken
country to the valley of the Kara Su.
Söğüt thus has a curiously dual character: it is a remote mountain town

reached with difficulty, but at the same time stands on one of the main
routes of communication in Anatolia. At present, since the construction of
the Anatolian railway at the end of the last century, the routes have shifted:
traffic to Eskişehır passes through Bilecik and Bozüyük, and Söğüt is more
isolated than ever. Yet this isolation was plainly not the case in a long period
which can be traced back to the sixteenth century. Evidence for an earlier
age is scattered but sufficient to give some elements of the situation in the
time of Osman.
The most famous monument of Söğüt is the tomb of Ertuğrul Gazi,
which stands outside the town beside the road as it rises toward Bilecik.
Although it has occupied this site since the fifteenth century at least, the
present structure is of the nineteenth and shows no trace of early Ottoman
work. Evliya Çelebi reports that the tomb was destroyed by Tamerlane.17 If
this is so, not only was the tomb extant in 1402, but the route through the
town was then in use. Although Söğüt is not mentioned in the narratives of
Tamerlane’s campaign, it is highly probable that it would have been attacked
and destroyed in the aftermath of the battle of Ankara when detachments of
the victorious army spread devastation throughout western Asia Minor. In
particular, they attacked the centers of Ottoman power, Bursa, Yenişehir,
and Iznik; destruction of the tomb of their enemy’s ancestor would have had
powerful symbolic value. A force could have been dispatched for this pur-
pose from Eskişehir, through which Tamerlane would have passed on his
way from Sivrihisar to Kütahya.18
The earliest mention of Söğüt is in the epic poem of Ahmedi (c.1334–1412);
it first appears in a Byzantine source in the history of Chalcocondyles, who
wrote about 1465, but remains on the site are far older.19 Although neither
buildings nor inscriptions appear to have survived from pre-Ottoman times,
scattered reused stones, including some evidently from Christian buildings,
indicate occupation in the Roman and late antique periods. The ancient
name of the site, however, is unknown.20 Traces of medieval settlement in
town and region are typically sparse, yet the sole noted remains, those of a
church in the yayla of Günyarık, may be of some significance (Fig. 1.4).
Nothing was standing in 1981, when the whole structure had been dug out
and much of it carted away for building materials, but the ground plan could
be reconstructed and there was sufficient evidence to suggest a chronology.
Fragments of columns, a chancel-screen, and sculptural decoration indicated
two phases, the first in Late Antiquity, probably the sixth century and the
other much later in the Byzantine period, most likely the thirteenth
Fig. 1.4 Byzantine remains at Günyarık
century.21 This thus appears to be a structure rebuilt by the Lascarids and

therefore a sign of Byzantine power in the area at a time immediately prior
to the arrival of the Ottomans. Medieval evidence for the route is more sub-
stantial, especially its most famous users, the forces of the First Crusade,
who, victorious at Nicaea, proceeded toward the central plateau and the
high road to the Holy Land. When they reached Leucae on the Sangarius,
which until recently preserved its name as Lefke, they divided their large
army into two groups for the march to Dorylaeum. These would have
followed the two natural routes between those places, one via Söğüt, the
other through Bozüyük.22 Söğüt was thus on a major route in the eleventh
century, and again, apparently in the fifteenth. The situation could hardly
have been much different in the time of Osman. The implication of this
geography—a town seemingly isolated in rough mountain country yet on
main communications—will be considered next.
The Summer Pastures
The sites of the summer pastures of Ertuğrul, which may be identified with
some certainty, raise some problems. Some were at Ermeni Beli, a pass in
the neighborhood of Ermeni Pazar, now called Pazarcık. This town lies at an
altitude of 800 meters in a long and narrow fertile plain enclosed by moun-
tains on the north and south. The open rolling country leads gradually
westward to a pass of over 1000m, then drops steeply to the broad plain of
İnegöl, some seven hundred meters below. To the east, the road from
Pazarcık follows a short and steep drop of about 150m to the valley of the
Karasu, then rises through that long gorge to the plain of Bozüyük.23
Ermeni Beli is usually identified with the pass along the Karasu, the route
between Bilecik and Bozüyük; it was known as Ermeni Derventi in later
Ottoman times, and was perhaps the site of the Byzantine Armenokastron,
which stood on the road somewhere in the vicinity.24 The earliest sources,
however, place Ermeni Beli near İnegöl, making it the higher and longer
pass which leads into that plain. In any case, the general area is the same,
the pass certainly existed, and had some importance in the late medieval
and early modern periods. Armenocastron has not been located, but the
natives of Pazarcık knew of a castle in the vicinity of the town. Unfortunately,
in spite of what seemed like specific directions and much searching in the
wooded hills to the south, it was not found.
Pazarcık itself contains nothing of the period but, like Söğüt, owes its
importance to a situation on a major trade route, until recently the main
highway between Bursa and the plateau. Its use is attested as early as the
sixteenth century; as the natural east–west route, it is probably of great
antiquity.25 In the time of the first European travelers, the main Anatolian
highway followed an alternate route south from Nicaea to Yenişehir and
Akbıyık to drop into the valley of Pazarcık; it was later replaced by the route
through Söğüt.26 Among others, Busbeck and Dernschwam followed and
described it in 1555 on their embassy to the Sultan: Dernschwam notes the
wooded hills with bad roads and wild boar before Ermen Pazarcık, and the
find region with good soil around the town whose prosperity was then
revealed by a mosque and a caravansaray, as well as the frequent trains of
camels and donkeys bearing grain to Bursa; from there, the road crossed
streams and mountains to pass through a narrow valley and oak forests
before entering the broader cultivated and sheep-raising plains of Bozüyük.
The most detailed account, that of Karl Humann, who passed through here
on his way from Bursa to the East in 1882, suggests that little had changed
in three centuries, his party climbed for three hours from the plain of the
İnegöl, gradually entering a forest of tall oaks and beech, supposed to be the
abode of wild deer, jackals, bears and panthers. After journeying for over
two hours and passing two derbents, or police posts, they reached the foot
of the mountain and the end of the forest. They saw three small sites with
ancient ruins before reaching Bazarcik then, seven kilometers after the
town, dropped suddenly to the valley gorge of the Karasu, which they
crossed by a bridge before finally reaching Bozüyük. The wild and romantic
scenery of the Karasu gorge (the Ottoman Ermeni Derbent) which the road
and railway now follow has long been admired and often described.27
Once again, a site which may be identified with some probability is situ-
ated on major routes, attested, as usual, only for later times, but inhabited by
the Romans and apparently of some significance in the Middle Ages. When
the armies of the second Crusade left Nicaea, they had a choice of three
routes to their goal: one, on the left leading to Dorylaeum, was short and
direct but dangerous; the middle route was safer and longer, but poorer;
that on the right, along the west coast, was longest but safest and best sup-
plied. The first was evidently the direct route through Söğüt, and the second
most probably the western road past Yenişehir and Pazarcık.28 In the event,
the Germans followed the first with predicted disastrous results; the French,
who went along the coast, did no better.
The actual pastures remain to be defined. If the text is to be taken liter-
ally, and the men of Osman considered as possessing large flocks of sheep,
they would have found grazing land in the fertile valley bottom rather than
in the steep and heavily wooded slopes on all sides. The first summer
pasture, therefore, is another place in rough mountain country, but on
major routes.
The other yayla was on the Domaniç Dağ, the massive chain which forms
the eastern extension of Ulu Dağ, the Mysian Olympus. The mountain tow-
ers above the plain of İnegöl from which foothills rise gradually until the
beginning of the steep main range; it has several distinct characteristics
(Figs. 1.5, 1.6). As the modern road leaves the plain, it climbs slowly and
crosses a ridge to the town of Tahtaköprü where the real ascent begins.
Below Tahtaköprü, and particularly to the east, stretch fertile valleys, yaylas
enclosed by hills, full of grass in the summer, and dotted with sheep. To the
south, a steep and rough road climbs the mountain soon to enter a vast and
dark forest, a dense growth of beeches unparalleled (in my experience) in
western Turkey (Fig. 1.7). The slow progress along the road probably exag-
gerates the size of the forest, but it is hardly less than ten kilometers across
and stretches as far as the eye can see east and west along the ridge. The
trees are so thick and tall that viewpoints are rare until the summit is
crossed and the treeless plains of Phrygia appear below at what seems a vast
distance. Ertuğrul certainly did not pasture his sheep in this forest or any-
where near the summit of the mountain.
Osman is supposed to have crossed the plain of İnegöl, evidently after
descending the Ermeni Beli, to reach this yayla. This is a far better route
than the rough mountain tracks he would have had to follow to reach the
area in a straight line from his home, and would lead him to quite suitable
pastures on the lower slopes around Tahtaköprü, far below the forest belt.
Sheep are in evidence there now, as they were when Browne passed by on
June 30, 1802. After leaving the village of Ortakoy at the southern edge of
Fig. 1.5 The Domaniç mountains, with some grazing ground [5]
Fig. 1.6 Looking down on the plain of İnegöl from the Domaniç range
Fig. 1.7 The great forest near Domaniç

the plain, he noted: “The road here lay through a wood where I observed the
largest flocks of sheep I have ever seen in Turkey. The shepherds told me
that the number here collected exceeded seven thousand: they were driven
to this spot for the advantage of being sheltered by the trees from the sun.”29
A more vivid confirmation of the tradition could hardly be expected.
Browne saw the sheep at an altitude of about 350m; the pastures of
Tahtaköprü are around 550m; the forests stretch from about 1000m to the
summit which in places exceeded 1900m. The significance of these figures
will become apparent soon.
This region and its forest are known from many reports of travelers
because the direct route from Bursa to Kütahya has crossed Mt. Domaniç
south of İnegöl for many centuries. It is first described by Bertrandon de la
Broquiere who commented on the height and length of the passage in
1432.30 The route in fact appears to be much older, apparently used by
Byzantine armies in the twelfth century, if not earlier. Manuel Comnenus in
1146 followed the steep and overgrown path at night by torchlight, moving
against the Turks; but the accounts of this and other campaigns are excep-
tionally unclear.31 The most detailed descriptions are later: Belon in 1555
commented on the long ascent, the abundance of tragacanth, the forest of
pine oak and beech, the snow, and wild boars; in 1645, Evliya Çelebi had, or
expected to have, far more trouble. The Turcoman guides he got from a vil-
lage six hours from İnegöl disappeared as soon as he entered the forest, and
the party proceeded, weapons drawn and on the watch, “in the midst of a
thousand fears” from the bandits who infested the road, until they reached
the villages of Domaniç.32
Trouble came more often from the state of the road and the weather than
from bandits; the sizes of the forest added to real or imagined problems. For
Lucas at the beginning of the eighteenth century, the forest was vast and the
road so poor that he was barely able to advance; Olivier, at the end of it,
crossing by night like Manuel Comnenus, reached daylight and the summit
in a thick forest of beeches, some, he said, a hundred feet high and three feet
thick.33 In 1814, Kinneir made his way up through thick snow and a forest
of pines, oaks, and beech until, “worn out with cold and fatigue” he reached
the lower slopes above İnegöl. Texier described the route in 1835 as passing
through fine forests of beech, oak, and chestnut, some with trunks twenty
meters high and two meters thick. Deep in the forest, the road gave out, and
he followed a stream bed through vines and thorns, with no sound but the
dull noise of the stream mixing like an echo with the rustling of the leaves.
In his day, the forest stretched twenty-four kilometers along the road.
Mordtmann almost got lost in 1852, but soon found that the pass was less
fearsome than he had been led to expect. For him, the beeches of Schleswig-
Holstein were mere dwarves compared to these; like his predecessors, he
spent many hours passing through the beeches before they mixed on the
descent with oaks, then planes.34
These accounts apparently describe a route like, if not identical with, the
modern road. Texier described another, the route from Bursa and İnegöl to
Tavşanlı and Aezani west of Kütahya. This was exceptionally tiring for the
horses, but passed through a country with magnificent woods, and splendid
views. Keppel followed it in the reverse direction a few years earlier, march-
ing through the forest on beaten snow; he noted that the oaks and beeches
were used for shipbuilding. Only the intrepid MacFarlane, coming from
Bursa in 1847, chose a somewhat different passage. An hour from İnegöl, he
passed a massive ruined caravansaray of brick and tiles, arriving soon after
at the foot of green hills and “a beautiful wild valley abounding with the
finest pasture,” but offering no sign of cultivation or human habitation. At
the village of Musal, he learned the dangers of living near the forest as the
villagers explained how their ox teams were pressed into service to drag out
huge trees to be used for the sultan’s ships. He continued from their past a
grand cliff of red rock, still a notable landmark, through a forest full of wild
beasts; “the overhanging branches of the trees and the dense foliage shut out
the sun and made a solemn gloom,” which cleared in time for the trip to be
made more spectacular by a solar eclipse. Soon after, they arrived at the hot
springs (of Oylat) where they found columns of steam rising from the
ground, basins of hot water, and a ruined stone bath building, “probably of
the Lower Empire.” From there, after an even thicker forest of beeches and
pines, they reached “the horrible bridle-path,” which led them down the
south slopes.35
The travelers reveal the essential features of the landscape: the plain of
İnegöl, the lower slopes and valleys suitable for pasture or cultivation, and
the vast beech forest, evidently of immemorial antiquity. The trees effec-
tively block communication over the mountain for all but small parties, and
even now the road is very bad, steep, and slow. The yaylas of Ertuğrul would
thus not have been high on the mountain, but on the lower slopes in the
vicinity of Tahtaköprü. Its pastures, like those of the Ermeni Beli, stand
astride a major route, and raise a similar problem about the tradition.
It is evident so far that the first places mentioned all existed; towns and
pastures are in the right places; the setting is authentic. The sources make
no mention of the roads, perhaps because they were not of interest for the
subject which deals, or seems to deal, with a heroic and nomadic age.
Nomads certainly need summer and winter pastures, and so the areas have
been defined. Yet a visit, or even careful study of a map, raises a serious
problem: which are the winter pastures, and which the summer? Söğüt, the
home, has an altitude of 656m; Ermeni Pazar is at 800m, and Tahtaköprü at
560, with pastures somewhat higher and lower.36 In other words, there is no
significant difference in elevation, or climate, between the three places; they
are equally hot in summer and cold in winter. It is hard to understand why
Osman should have moved his sheep between them and, if he did, what
advantage he would have gained. Even more curious in this respect is the
geography of the land around Söğüt. There is a good yayla in the mountains
to the south, and winter pastures could have been found, if they were avail-
able, in the valley of the Sangarius (altitude 180m) or in the land to the
northwest toward Vezirhan. The places fit the historical tradition perfectly
well, but not its nomadic component.
First Battles and Conquests (APZ caps. 3, 5)
As Osman and his followers passed from their winter to their summer pas-
tures, they crossed the plain of İnegöl which lay on the direct route. When
the tekfur, or Christian commander, of İnegöl (his name was Aya Nikola)
harassed these peaceful movements, Osman appealed to the tekfur of
Bilecik, and arranged to leave his heavy goods and valuables in the security
of his castle during the seasonal movements.37 Osman planned to set fire to
İnegöl by night, but Aya Nikola laid an ambush at the end of the Ermeni
Beli. Osman, however, made a successful attack. In the ensuing battle, his
nephew Bay Hoca was killed; he was buried near the village of Hamza Beg
where a ruined caravansaray stood near his tomb.
This account is consistent with the traditional image of a nomad tribe
migrating between pastures, alien to town life, without a defensible base,
and thus in need of the borrowed security of Bilecik. The arrangement with
that tekfur is typical of many stories which show the good relations between
Osman and the Christian population; the battle of Ermeni Beli was the first
of many hostile but successful encounters with his Byzantine neighbors.
The places mentioned are real, most of them attested for the period.
İNEGÖL is now a large and thriving town famous for its meatballs, the
İnegöl köftesi. It occupies the center of a rich, cultivated plain and has long
been of importance. The Ottoman tradition suggests that it was a Byzantine
base whose defenses were no doubt intended to protect the plain and the
routes through it. These, as noted, were the main road from Bursa to
Eskişehir, and the western of the roads between Nicaea and the plateau. For
the early history of the town, which is totally obscure, this text gives a valuable
and circumstantial piece of evidence: its Byzantine name, St. Nicholas. Aya
Nikola, the name of the tekfur, plainly represents a place name, and would in
fact be highly inappropriate for an individual. The modern name İnegöl is a
modification of the Ottoman Aynegöl, the form which invariably appears in
early sources. In this case, a Byzantine Ayos Nikolaos was simply Turkicised
to become Aynegöl; the normally accepted identification of the town with a
Byzantine Angelokome is based on a fancied resemblance of name.38
İnegöl contains no castle or remains of great antiquity, but a mosque and
bath attributed to Beyazit I indicates some importance at the end of the
fourteenth century, while the great complex of buildings of Ishak Pasha
with its splendid mosque certainly indicates considerable prosperity, much
of it no doubt from commerce, in the middle of the fifteenth. It was a flour-
ishing town in the time of Evliya Çelebi.39
In the past, the surroundings of the town had a rather different aspect:
the edges and part of the plain were heavily wooded, but there was sufficient
cultivated land to justify its description as “a country as much favoured by
the bounties of nature as it is cursed by the oppression of man”; the center,
however, was treeless and swampy where the river flowed through low-lying
parts. In the late nineteenth century, there was extensive agriculture in the
plain, aided by stone embankments which contained the extensive
marshes.40 There is no reason to doubt that the town and plain were well
worth fighting over in the time of Osman.
Osman’s nephew fell in the battle and was buried near the village of
HAMZA BEY. This, too, is a real place, situated at the northern edge of the
plain where the road enters the pass which leads toward the plain of
Yenişehir. Mordtmann reports that it had a bridge where the road crossed
the Koca Çay and a fine mosque, probably the signs of prosperity from the
transit trade.41 The ruined caravansaray supposedly near the tomb is a sig-
nificant detail, too circumstantial and easy to check to be a fabrication. If
such a structure were already ruined in the late fifteenth century, it would
have functioned at an earlier, but already Ottoman, time (the Byzantines
did not construct such buildings) and the importance of the place be
brought back much closer to the period in question.
BILECIK, whose location has always been known, is the first site which
actually contains remains from the appropriate time. The old town stands,
not very comfortably, on rocks and steep slopes at the base of a high ridge of
the edge of a valley which leads down to the moderately broad plain around
the Kara Su (Fig. 1.8). Streets are rough and steep, and the rocks seemingly
so dangerous that they were fastened to the cliffs by chains, to keep them
from crushing houses, in the mid-nineteenth century.42 The modern visitor
passes through the new town, a dull place built along the highway at the top
of the ridge, but can easily see the old part from above as he leaves the town
toward the south.
The last hills at the bottom of the ridge contain the most important mon-
uments so far considered: the fortress and the mosque of Orhan. The lord of
Bilecik had a castle in which he could protect the goods of Osman; badly
ruined scraps of its walls still stand on an isolated hill about the mosque of
Orhan. They are built in coursed rubble with intermittent bands of reused
brick, techniques which could indicate a late Byzantine origin—perhaps a
poor or provincial work of the Lascarids, or Palaeologans.43 In any case,
they proclaim the town as one existing when Osman came to the region.44
Here, then, is the first structure mentioned in the sources which can be
identified with standing remains. If the town contained nothing else, it
would be plausible that the stories were inspired by the ruined castle. The
mosque of Orhan, however, leaves no doubt that the place was an early
Ottoman settlement, and one of considerable importance (Fig. 1.9). The
Fig. 1.8 The site of Bilecik

Fig. 1.9 Bilecik: Orhan’s mosque
building has been well studied, its style and techniques of construction
analyzed; there is no doubt that it is a product, and a quite distinguished
one, of the mid-fourteenth century.45 The town also contains a well-built
imaret plausibly attributed to Orhan.46 In this case, the remains combine to
confirm the narrative of the sources.
“1286” First Victory (APZ cap. 5)
In the following year, Osman made his first strike outside his immediate
homeland—against his enemy the tekfur of İnegöl.47 He attacked and
burned the small fortress of Kulaca and slaughtered its garrison. This marks
the beginning of his career of conquest, and the occasion for increasing
concern among the Byzantines.
A place called KULACA is still on the maps and very easy to find, lying
not far off the main Bursa-Eskişehir highway, its location marked by a large
signpost to the source of local renown, a factory for tomato sauce. The town
lies on the banks of a stream about five kilometers east of İnegöl in a site
good for agriculture but difficult to defend. Its easy accessibility has not kept
it from remaining unknown: the friendly and helpful villagers knew of no
other foreigners who had come to see the ruins. Long before living memory,
however, Carl Humann passed through in l882 and revealed the importance
of the site: it lay on the main route from Bursa and İnegöl to the valley of
Bazarcık and thence to the plateau. In his day, as now, the place had a good
bridge, a feature whose value becomes evident from his account of travers-
ing this marshy country, crossing with enormous difficulty streams swollen
by the spring rains. The site would thus be of some real value for controlling
the plain and communication through it, which, to judge by Humann’s
mention of a ruined stone caravansaray at Kurşunlu, just east of Kulaca, fol-
lowed this route in the Ottoman period. E. H. Ayverdi has investigated the
place for its historical remains, and concluded that it contained nothing of
architectural interest.48 In general terms, his judgment was confirmed by
the visit: although the locals knew that their town was the first conquest of
Osman, they could show nothing from his time, expressing only the rather
wishful suggestion that their old mosque (perhaps of the nineteenth cen-
tury) somehow represented or replaced Kulaca Hisar.49 This seemed like a
case of a schoolmaster’s fancy identifying a modern site with one of an
heroic age, yet the village had the right name and occupied the right loca-
tion. A small bit of evidence came from a ruined bath whose architecture
seemed to evidence no great age; it contained a reused Byzantine capital,
probably of the sixth century. This single stone cannot demonstrate the
antiquity of the site—it could have been brought from elsewhere—but it at
least raised the possibility of Byzantine settlement there, and with it indirect
and dim confirmation of the chronicles.
“1286” (APZ cap. 5)
The local tekfurs, now alarmed, formed their first alliance against Osman,
and their defeat was his first victory over a united opposition. The com-
manders of İnegöl and Karacahisar, aided by Kalanos, brother of the latter,
joined forces against the new threat. Osman advanced to a place variously
reported as Ikizce, Ekinci, or something similar, but actually to be read,
with the slightest emendation, as Alınca.50 This lies between İnegöl and
Pazaryeri. There was a great battle here where the one crossed the Domaniç
Beli. The Ottomans, naturally, were victorious, but Saru Yatı, brother of
Osman, fell, at a place where there was a great pine tree, called the Tree with
Lights (Kandilli Çam) because of a light that shone there on many occa-
sions. He was buried beside his father in Söğüt. Kalanos also fell; Osman
ordered his men to rip open his belly and to scratch at the ground like a dog
and bury him at a place thereafter named It Eseni (“Dog Scratching”).51
If the mention of the Domaniç pass has any significance, though, and
if the infidel commanders met halfway between their respective bases,
Alnca and It Eseni would be in the right location. Before Alınca was
brought into the discussion, I supposed that a site with the possibly
significant name of Kandilli might be considered. Although that is now
unlikely, I leave the description of the site as another example of the local
fortifications.
The village of KANDILLI stands on the Sari Su at the southern edge of a
broad plain, west of İnönü, and almost due south of Bozüyük. The Kara Su
rises nearby, the village being situated near the watershed between Bithynia
and the broad Phrygian plains. Kandilli lies at the foot of a long oval hill
which rises rather steeply over the adjacent plain. A gorge separates it from
the mountain mass on the south which bounds the broad and fertile plain
of İnönü; this comes to an end just to the west in the barrier of the Kandilli
Dağ, an offshoot of the main Domaniç range. The village occupies a poten-
tially strategic location with easy routes north along the Kara Su, west to
İnönü and Eskişehir, and to the south a track which leads over hills and
mountain pastures to Kütahya.52
The hill above the village offers splendid views over the surrounding
plains and mountains and contains remains of some interest.53 About half
the summit is, or rather was, surrounded by a wall invisible from below. The
only substantial part now standing consists of a tower on the south side at
the end of a cross-wall which protected the approach from the gentler west
slope. The tower is built of core of mortared rubble with a facing of well-cut
stone in regular courses, with a double band of brick. Most of the stones,
according to local shepherds were long since removed to build the mosque
in the nearby village of Karaağaç. The cross-wall displays two distinct kinds
of mortar, suggesting perhaps two periods of construction. These dilapi-
dated remains indicate a Byzantine origin—the stonework could as well be
Seljuk, but the brick bands are typically Byzantine—quite probably of the
Comnenian period when such a masonry was in common use.54 The rest of
the circuit has left little trace, but enough to show that the hilltop was sur-
rounded by a wall which enclosed an area of about 400 × 100 meters. Its
large size and location at the edge of fertile lands suggest an origin as a ref-
uge sit of the Dark Ages, the seventh to ninth centuries. Whatever its exact
date, it is evident that a Byzantine fortress stood here which in its last stage
probably represented parts of the Comnenian effort to retain control of
their eastern border.55 The fortress would no doubt have been standing and
in good condition in the time of Osman.
Karacahisar, Sultanönü, and Conflict

with Germiyan: “1288–1300”
The conquest of Karacahisar in 1288 looms large in Ottoman tradition.

Here, the first Friday prayers were held in the name of Osman; the town was
thus the scene of his proclamation of independence from the Seljuks, then
in fatal decline. The confusing narrative of the events and attendant circum-
stances indicate that a group of legends sprung up around these conquests
to reflect the importance attached to the place. Since Karacahisar is associ-
ated with the acquisition of the whole region called Sultan Önü—which
includes İnönü and Eskişehir as well—it may be best to treat these places
together in an attempt to analyze the tradition and to understand its
significance.
APZ (cap. 2) reports that the Seljuk sultan Ala ad-Din settled Ertuğrul in
the region between Bilecik and (Afyon) Karahisar that had been plagued
by attacks of the Çavdar Tatars; at that time the tekfur of Sultanönü and
Karacahisar was paying tribute to the sultan. (According to the Anonymous
[p. 11f.] the tekfurs of Karahisar and Bilecik were tributary to the sultan.)
Neşri attributes the conquest of Karacahisar to Ertuğrul, a pious act to
assure that proper respect was paid to the Seljuk Sultan. On the death of the
Sultan, however, the Muslims are supposed to have lost control of the
town.56 A romantic tale of love and war purports to recount events affecting
the whole district in the days of Ertuğrul, at a time when the young Osman
was still unsuccessful in his suit for the hand of Malhun, daughter of the
learned Sheikh Edebali of İt Burnu. In this, the beys of Sultan Öyüğü and
Eskişehir appear in alliance with the Christian Mihal of Harmankaya to
attack the bey of İnönü, friend and protector of Osman. The main action is
set around the ruined castle of Inhisar in the district of İnönü, where Osman
defeats the bey of Eskişehir, his rival for the affections of Malhun, and gains
the alliance of Mihal and of the Sheikh.57
The confused geography of this account—Sultan Öyüğü and Eskişehir
are really the same place; Inhisar is in the valley of the Sangarius far from
İnönü, sharing only a resemblance of name—reveal it as a folktale, appar-
ently of no value. When that story is taken together with another passage
from the same source, however, a certain historical reality seems to emerge.
According to Neşri, when Osman came to power, deputies of the Seljuk Ala
ad-Din II were ruling in Eskişehir of (the province of) Sultanöyüğü and
İnönü, while the contemporary Haci Bektaş mentions Ermeni Beli as mark-
ing the frontier between the Byzantine and Seljuk realms.58 This may reflect
a real situation in which the frontier region at the northwestern edge of the
Anatolian plateau was still under the control of Konya, though perhaps, if
one may judge from the independence of action attributed to these beys, in
a somewhat attenuated form. The tradition, in any case, shows these rulers—
the Christian tekfurs of Bilecik and Karacahisar and the Muslim beys of
Eskişehir and Inönü as independent of Osman at the beginning of his career.
More serious tradition begins with the conquest of Karacahisar in 1288,
an event accompanied by suitably edifying circumstances. When Sultan Ala
ad-Din heard of the victory of Osman over the infidel coalition, he gathered
a large army to attack Karacahisar whose tekfur, always hostile (both to him
and to Osman: he had participated in the joint attack of 1286) was allied
with the emir of Germiyan. Osman’s forces joined the enterprise but the
Sultan had to withdraw because of a Tartar attack, leaving Osman to carry
on with his blessing. After a few days, the castle was stormed and con-
quered, its commander captured, and its houses handed over to the follow-
ers of Osman who turned the place into a Muslim city. The Sultan, delighted
by the news, conferred on Osman the symbols of delegated authority: a
banner, a tent, horses, and weapons, thus granting official recognition to his
exploits.59 Osman thereafter followed a policy of good relations with his
Christian neighbors.60
The only problems came from Germiyan, constantly hostile to the
Ottomans. When Osman set up a market by the hot springs of Eskişehir,
many of the local Christians came, among them merchants of Bilecik who
brought drinking glasses of good quality. On one occasion, a man from
Germiyan took a glass from a Christian merchant without paying. Osman,
on hearing of this, chastised the Turk, ensuring justice for the Christians so
successfully that the market became large and flourishing. Karacahisar
gained importance in a different role: it apparently became a major fortress,
the place where Osman and Mihal returned after their expedition north of
the Sangarius in 1293.61
The tradition reports rather anomalously, however, that Karacahisar
remained empty for some time until people from Germiyan asked for
houses there; Osman granted them, and the place grew and prospered.
Many churches were turned into mosques, and a market established. When
the people asked that a Friday prayer be organized and a kadi (religious
judge) appointed—signs of city life and of independence—Osman took the
bold step of having the prayer read in his own name, thus proclaiming his
independence in 1299.62 This date has been taken to mark the beginning of
the Ottoman empire. After organizing the administration of the city, Osman
turned the whole district, which was also called İnönü, over to his eldest son
Orhan to rule.
Within a few years, he was called on to defend his province when the
Çavdar Tatars, inhabitants of the region of Germiyan, attacked the market
of Karacahisar.63 Orhan at that moment was having his horses shod at
Eskişehir. He advanced rapidly, met the Tatars at a ruined castle called
Oynaş in the mountains defeated them, and brought back the stolen goods
with numerous captives to Karacahisar. There, Osman ordered the Tatars,
who were fellow Muslims, released and made a peace with them which
lasted until the end of the fourteenth century.64
Everything in the sources indicates that Karacahisar and its neighbors
were of considerable significance, for both real and symbolic reasons, in the
rise of the Ottomans. All the places mentioned in these narratives may be
identified, and most of them visited. The results of such investigation, com-
bined with evidence from other, earlier sources, raise serious questions
about the Ottoman traditions, and allow this part of them to be seen in a
quite different light from those which have so far been considered.
The location of KARACAHISAR is well established, and its site easy to
reach (Fig. 1.10). The castle stands on a commanding ridge overlooking the
narrow valley of the Porsuk, on the south side of the river about ten kilometers
southwest of Eskişehir. It was plainly designed to control the passage along
Fig. 1.10 Karacahisar barely visible on top of the hill [10]

the river between Eskişehir and Kütahya or, more appropriately, Dorylaeum
and Cotyaeum. The town associated with it, Karacaşehir, lies slightly down-
stream at the confluence with a tributary. The center seems to have shifted
here at a fairly early date, for a document of the late fifteenth century shows
the town as the headquarters of a kaza, or administrative district, a distinc-
tion it still retained two centuries later. The document mentions a zaviye
(dervish lodge) in the town, and the names of various dependent villages.
Dernschwam in 1555 learned that Karacaşehir had a castle and a population
of Turks and Armenians, but its market was empty; since it lay off his route,
he did not visit it. Yet for Haji Kalfa, the place was a fine town with good air,
north of Kütahya, at a distance of four hours from İnönü, and with a castle
conquered by Osman.65 Subsequently, it became a village and attracted little
attention until the late nineteenth century, when scholars attempted to
define the site of ancient Dorylaeum. At one time, it was proposed that the
earliest settlement of Dorylaeum lay here; General von der Goltz, told that
the ancient name of the place was Dorila (on dubious authority, it seems),
visited the site and found that it contained no trace of antiquity, a defect
which he modestly attributed to his own ignorance of archaeology. In fact,
Dorylaeum was soon firmly located at Eskişehir, and Karacahisar recog-
nized as a site from the Middle Ages, although subsequently some late
antique as well as Byzantine inscriptions were discovered there.66
Karacahisar is a large castle in a dominating position, formerly impossi-
ble to visit because it has become the site of a radar base, but thanks to its
importance in Ottoman history and his reputation as the pre-eminent his-
torian of the Ottomans, Prof. Halil Inalcık got permission to investigate the
site.67 Descriptions and photographs of the fort reveal its significant charac-
teristics.68 Its walls were constructed of a core or mortared rubble faced
with regularly coursed fieldstones, among them some courses of long flat
stones and an occasional band of three or more bricks. The masonry con-
tained a great many reinforcing timbers. The use of the flat stones and brick
are Byzantine characteristics; the irregular masonry suggests a late date,
perhaps in the twelfth century. The castle is probably to be identified with
one of the unnamed forts in the region of Dorylaeum which Manuel
Comnenus provisioned in 1175 prior to rebuilding Dorylaeum.69 Remains
in the town are modern or at best old structures rebuilt. There is thus no
trace of the mosque which a document of questionable authenticity attri-
butes to “Sultan Osman Han Gazi.”70
If the castle of Karacahisar were built some hundred years before the rise of
Osman, it was no doubt still standing and maintained in his time. It would
have been a valuable frontier defense for the Seljuks, who controlled the
region by the late twelfth century, and even more significant for their suc-
cessors, the emirs of Germiyan. The castle controls the main routes from their
capital, Kütahya, to Eskişehir, Ankara and the central plateau; if it were in
enemy hands, the approaches to the heartland of their state would have been
under serious threat. Its history in this period, however, is totally obscure.
The tradition is curiously ambiguous about the role of Germiyan at
Karacahisar. Leaving aside stories of the heroic days of Ertuğrul, the sources
show that the place was a center of enmity between Osman and Germiyan,
and sometimes allow the local power of the latter to be seen. As already
noted, Osman is supposed to have taken it from infidels when Germiyan
was negligent in attacking them. Osman secured justice for the Christians
against Germiyan whose men are portrayed as mere visitors to the market at
Eskişehir; later, but before 1299, the surprisingly empty town was populated
by settlers from Germiyan. The most curious incident is a disingenuous one
which illustrates the pristine virtues of the simple tribesman, Osman: after
Karacahisar had been organized as a city, a man from Germiyan came,
seeking to buy the right to collect taxes on the local market. When he was
brought before Osman, the ruler was completely ignorant of taxes, reacted
with some indignation when they were explained to him and drove the man
from his presence. Finally, however, when he learned that taxes were the
normal rule everywhere, he allowed a small percentage to be collected.71
Then, as has been seen, he attributed the local administration to Orhan who
had to fight against Tatars from Germiyan.
These stories suggest a serious problem in the tradition, somewhat simi-
lar to the ambiguity in early Roman history about the Sabines or Etruscans,
foreign people whose major, even dominant role, could not be concealed.
The Ottoman tradition no doubt contains a similar distortion. The early
chroniclers plainly wanted to magnify the importance of the conquest of
Karacahisar, and thus the establishment of Ottoman power on the plateau;
hence its retrojection to the reign of Ertuğrul, and its association with the
Seljuk Sultan. Yet at the same time, they could not escape the known fact
that Germiyan was powerfully involved here: by their own admission, the
place had a Germiyanid population, and someone from Germiyan could
propose to collect taxes there. It seems most probable that the tradition is
attempting to cover the unwelcome fact that this strategic place near the
Ottoman homeland was in the hands of Germiyan long after the supposed
date of the Ottoman conquest. This, indeed, seems an inevitable conclusion
from other evidence. An inscription of 1300 shows the ruler of Germiyan in
control of Ankara, under the real or nominal suzerainty of Ala ad-Din

Kayqubad III.72 Although their power over Ankara seems to have had a
brief life—the city was apparently subject to the Mongols by 1304. It is diffi-
cult to believe that such a great extension of empire could have taken place
when a crucial fort on the road between Kütahya and Ankara was in the
hands off their Ottoman enemies. The Ottoman conquest thus needs to be
placed after the first years of the fourteenth century; how much later will be
considered next.
Orhan defeated the Tatars at a ruined mountain castle OYNAŞ, a real
place, discovered by the indefatigable Major von Diest in 1896. The castle
stands in the rough country of the “highlands of Phrygia,” a remote region
between Eskişehir and Afyon famous for its ancient monuments. It is about
fifty kilometers south of Eskişehir, in an area far different from any so far
considered, full of easily defensible hills and mountains, narrow valleys and
small plains, and extensive forests. The castle, in modern times called Asar
Kale, occupies a steep round-topped hill overlooking a broad space which,
according to von Diest, was clearly marked with lines resembling a network
of streets. This lower area preserved the name Oyneş, and seems to repre-
sent the ancient metropolis of Phrygia. Later investigations of the region
have revealed that the fortress was in origin Phrygian, but was reoccupied
in the Middle Ages, perhaps by the Byzantines or somewhat later. In the
plain below are fragments of a medieval Byzantine church and a Seljuk
tomb, both witnesses to the existence, if not the importance, of the place in
the late Middle Ages. The location, although remote, has the advantage of
dominating the road which leads south from Eskişehir and Seyitgazi to
Afyon, towns of some importance under Byzantines and Seljuks.73 Although
real, Oynaş is exceptionally far from the area of activity which even a gener-
ous tradition might attribute to Osman. There seems no reason to believe
that his son would have fought a battle there as an independent agent, far
from his homeland. Three possibilities seem open: either the identification
is wrong; or the tradition is correct, and Orhan participated in a battle here,
in which case his detachment is probably to be seen as marching in support
of another power and perhaps under its leadership; or Oynaş has been
brought into the record because there is a place called Karacaşehir which
contains another castle about five kilometers southwest of it.74 In that case,
the identity of name between this and the settlement below Karacahisar
could have produced some confusion and attribution of action by Osman to
this far-away region. The case is quite different from those of İnegöl or
Bilecik, say, where a real place is in a location otherwise plausible for the
events described. There seems little reason to believe that the young Orhan
would have fought a battle at Oynaş, far from his homeland, as an inde
pendent agent.
İNÖNÜ, the second center of Sultanönü, takes its name “Front of the
Cave” from the great caverns in the cliff directly behind the town. It, too,
appears in the saga of Ertuğrul’s days but, curiously, nowhere else in the
tradition. The site is so distinctive that its identity has never been in doubt;
it is described by several travelers, and contains monuments, not always
easy to interpret, of an appropriate period. The town lies at an elevation of
872 meters at the foot of the Dutluca Dağ, at the south end of the broad
plain of the Sarı Su, which extends from the vicinity of Kandilli, gradually
rising toward Eskişehir (Fig. 1.11). A low ridge of hills separates this plain
from that of Bozüyük, while an easy pass leads across the plateau to the val-
ley of the Porsuk. Since the town is conveniently situated on routes which
lead from the north and northwest southward to Kütahya or eastward to
Eskişehir and the central plateau, it has been frequently visited.
In the seventeenth century, Haji Kalfa described İnönü as having many
caves, one of them inhabited, and a castle with a garrison of a dozen sol-
diers, located on a mountain reached with great difficulty. He noted that the
town stood on roads which led to Akbiyik (north of İnegöl), Eskişehir, and
Fig. 1.11 The town of İnönü

Karacaşehir. Otter gave a suspiciously similar description in 1736, when he

passed the town without stopping on his way from Bilecik and Bozüyük to
Eskişehir.75 When General Koehler visited in 1800, he came from Kütahya
through “fine pasture lands mixed with good timber trees” until he looked
down on the broad plain of İnönü, which in his time was a large village
under a precipice with caverns, some the abode of eagles. “One enormous
cavern,” he reported, “is shut up in front by a wall with battlements and tow-
ers and seems to have served as a sort of citadel.” He passed from there to
Söğüt.76 Keppel arrived from the opposite direction in 1831, noting the
numerous Greek pillars in the graveyard and reporting the existence of a
castle halfway up the mountain, which he could not visit because the frost
had made the route too slippery.77 As he remarked, the reported castle may
have been no more than the walled up cave; no other fortification seems to
be known in the region.
The most striking landmark of İnönü is certainly the great cliff face rising
sheer above the town, and the huge cave in it.78 Most of the wall which
blocked the entrance has now disappeared, but surviving traces show a
masonry of squared stones without brick over a core of mortared rubble
which has been strengthened by an extensive network of wooden beams
(Fig. 1.12). In style, this most closely resembles the walls of Afyon Karahisar,
a Seljuk work of the thirteenth century. It appears in any case to be Seljuk
rather than Byzantine and may thus be taken as a monument of the period
immediately before Osman. Reused stones scattered through the town
extend the history back by providing evidence of a substantial settlement of
the Roman and late antique periods.
The most notable monument of the town, however, is the mosque of
Yadigar, situated just below the main cave. According to the inscription over
its door, this was the work of Hoja Yadigar, son of the Sultan Ali, in the year
771 (1369/70).79 Analysis of the architectural style has shown that the
mosque is similar to others of that date, thus confirming the authenticity of
the inscription.80 The name of Sultan Ali raises central questions: Who was
he and who ruled İnönü at the time? The answer is not at all obvious: there
is no Ottoman sultan named Ali, and the date is too late for reference to
Alisir, the ancestor of the Germiyans, to be intended. Yet it does appear—
though the identification is by no means certain—that Ali may have been a
byname of Suleymanshah, who ruled Germiyan from 1361 to 1387.81 If so,
İnönü would not have been Ottoman even at that late date. It is thus not
surprising that İnönü finds no place in the traditions about Osman. APZ
only mentions the sanjak of İnönü being entrusted by Orhan to his son
Fig. 1.12 İnönü’s fortified cave
Murad after the conquest of Bursa and before that of Iznik, with no indication
of when and how it came into his hands.82 İnönü only appears in history
with the inscription of Hoca Yadigar. It was quite probably the seat of a
small independent emirate.
The location of ESKIŞEHIR has never been in doubt, and its identifica-
tion with the ancient and Byzantine Dorylaeum is well established. It occu-
pies one of the most strategic sites in Anatolia, the point where the routes
which lead from Europe via the Bosporus, the Dardanelles, or the Sea of
Marmara, converge after passing through the broken country of Bithynia
and before diverging to the south and east through the central steppe to the
neighboring regions and countries (Fig. 1.13). The site has therefore long
been occupied by a city, which in the Middle Ages was a major military base
and in modern times has become a principal railway junction and indus-
trial center. It is far more important than any of the sites so far discussed.
Byzantine Dorylaeum was a great fortress, one of the posts where the
emperor and his troops stopped to gather reinforcements on their campaigns
Fig. 1.13 Looking east from Sultanönü
to the East. After the collapse of Asia Minor to the Turks in the late eleventh
century, it became the scene of the first great victories of the First Crusade,
whose forces necessarily followed the route which led here in 1097. Their
successors of the Second Crusade took the same road in 1147, but enjoyed
far less success. During most of the twelfth century, the town appears to
have been deserted and its region, given over to nomadism, served as a sort
of no- man’s-
land between Seljuks and Byzantines. Finally, Manuel
Comnenus decided to strengthen his border by refortifying Dorylaeum, an
act considered a provocation by the Turks, and which formed a prelude to
the disaster of Myriokephalon in 1176.83
The town fell almost immediately to the Seljuks, for al-Harawi, who vis-
ited it not long after 1177, described it as a place with hot springs on the
frontier of the infidels, and referred to it by its new Turkish name.84 The old
name Dorylaeum was forgotten by the Turks, who called the place Sultan
Önü, seemingly appropriate as meaning “In front of the Sultan” (that is, a
place on the border) but apparently derived from the alternative form
Sultan Öyüğü, “The Sultan’s Tumulus,” a toponym which well reflects the
nature of the site. The Byzantine fortifications were in fact built on a low
hill, an ancient tumulus, about two kilometers from the hot springs which
form the main attraction mentioned by ancient and modern sources. The
present city of Eskişehir is built along the river, incorporating the hot
springs; the fortress lies outside in a suburb now called Sar Hüyük, “The
City Mound.” The sources, however, do not apply different names to these
two parts of the settlement, but regularly include the hot springs and the
river in Sultan Öyüğü.
Ibn Said, writing in the middle of the thirteenth century, used the Arabic
form of the Turkish name, Sultanyuki, to denote the town which he
described as lying eight parasangs west of Ankara and containing baths
with natural springs of hot water. In his day, the land between the two cities
was well cultivated.85 The information at which he hints is given in remark-
able detail by slightly later sources of considerable importance for the
period immediately preceding that of Osman. The first of these is the
inscription of the minaret of the Alaeddin mosque in the center of the city
which proclaims that it was rebuilt by Jebrail ibn Jaja during the reign of
Kaikhusraw III (1265–1284) in a year which may apparently be read as 666
(1268).86 This shows that the city was in Seljuk (but really Mongol) hands
and evidently prospering when Ertuğrul and his tribesmen were settling in
the borderlands northwest of the city.
Far more significant, though, is the testament of the same governor,
whose name appears as Nureddin Jebrail son of Jaja. It has survived in full
to reveal a considerable amount about the city and its region.87 This document,
written in 1272, shows that the city and its region were in a flourishing state
under the Mongol governor. Although the document gives grandiose
titles—which fill five lines of text—to the insignificant Kaikhusraw III, then
about ten years old, it is accompanied by a text in Mongolian filled with the
name of witnesses including two successive governors, thus constituting a
validation of the will and revealing the texts as official documents of the
Ilkhanid government.88
The emir was concerned to set up an endowment for the mosque he had
built in Sultanyuki. To judge by the property attached to it, this must have
been a substantial foundation. In addition, he restored seventeen mosques
which had fallen into decay, as well as a mosque in a caravansaray—where
he made provision for the study of the Koran and a zaviye.89 The number
and variety of the establishments witness not only the wealth and generosity
of the donor but the size and prosperity of the city. Sultanyuki, it seems, had
at least seventeen private mosques (many of which might, of course, have
been quite small) in addition to the congregational mosque of Alaeddin
whose minaret the emir had restored. The caravansaray indicates substan-
tial commercial activity in a town which had evidently made a striking
recovery from the desolation of late Byzantine days during the century of
Seljuk rule. Of these monuments, only the congregational mosque has

survived, and that in a completely modernized state. In the earlier years of
the twentieth century, however, a bridge of apparently Seljuk origin still
stood over the Sarı Su near its confluence with the Porsuk.90
This source dates from the time of Ertuğrul; the city also appears in asso-
ciation with Osman: APZ uses the name Eskişehir and mentions it as a
place where Osman held a weekly market near the baths, where he had
occasion to chastise a man from Germiyan; where he held Bayram prayers;
and where he had his horses shod (APZ caps. 9, 14, 21). Finally, when
Osman died, he left behind, among his modest possessions, a herd of horses
at Sultanönü.91 None of these indicates that Osman actually ruled the place.
Sources from the next generation give less detailed but surprising informa-
tion about the city. Balban, al-Umari’s informant, narrates the energetic
campaigns of the Mongol governor, Timurtash son of Choban, to extend his
power over the Turcoman principalities of Anatolia in the 1320s. Among
his conquests was the territory of Sultanyuki, described as having no real
city, but sprawling villages and broad plains. The location of this land
between the domain of Suleyman Paşa (who ruled Kastamonu) and that of
Germiyan leave no doubt of the identification.92 The inevitable conclusion
is that Sultanyuki was the center of an independent state which occupied
the northwest corner of the plateau in the last years of Osman.93
The contemporary traveler, Ibn Battuta, who passed through Anatolia in
1335, did not visit the town, but he did meet two of its natives. In Nicaea, he
stayed in the house of the imam and jurist Alaeddin al-Sultanyuki, who
introduced him to the wife of Orhan and who had visited the holy places of
Arabia; and in Kastamonu he encountered the learned imam Shaykh
Tajeddin al-Sultanyuki who had studied in Iraq and Persia and had likewise
visited the holy cities.94 Both were evidently natives of Sultanyuki who had
settled abroad in the cities of the thriving neighboring principalities.
Another local, the molla Mahmud ibn Mehmed al-Sultanoyugi, achieved
even higher distinction. After studying Arabic science, the sacred law,
Koran exegesis, and the tradition with the learned men of his time, and
achieving distinction in all fields, he was appointed by Orhan to be kadi of
Bursa, a role he filled with considerable distinction for many years. Late in
life, Murat I sent him as head of an embassy to Germiyan. He is alleged to
have been kadi for forty years.95 The learning and accomplishments of these
men suggest that the city had a tradition of education, some of it available
locally—such as the Koran lessons in the mosque of the caravansaray—
most of it sought elsewhere, evidently by families who had the resources to
send their children to travel or study abroad. These scholars may have had a
special function of spreading orthodox high Islam to states and places that
didn’t yet have a system of Islamic education, like the Ottoman realm, whose
first medrese was opened in Iznik soon after its conquest by Orhan in
1331.96 Alternatively, the presence of these men away from home and the
ambiguous comment of al-Umari about the city may indicate that greater
opportunities were felt to exist in the larger adjacent states. Nevertheless, it
is clear that Sultanyuki was still independent and prospering to some degree
in the early years of Orhan. Here, for the first time, is direct contradiction of
the tradition by contemporary sources. Their account, of course, is to be
preferred, and Sultanyuki/Eskişehir seen as outside Ottoman territory at
least until the reign of Orhan.
By the late fifteenth century, when some details are known from a list of
vakf (endowment deeds) in the province, the town was known by its mod-
ern name, Sultan Öyüğü being reserved for the district; Şehir Hüyük, the
site of the Byzantine castle, also appears, apparently as a separate settlement.
The document mentions a zaviye in Eskişehir, different from that endowed
by ibn Jaja.97 The fate of Eskişehir during the next three centuries is obscure.
In the late seventeenth century, it was large and prosperous, according to
Evliya Çelebi. It was the capital of a kaza, the seat of various officials, and
contained a ruined castle built by the “tekfur of Bursa” and captured by
Osman from the Byzantines in 1331. The town consisted of seventeen
mahalles, or districts, and had many mosques, some medreses (but these
were not built of stone), seven each of children’s schools, tekkes, and cara-
vansarays, and a market with eight hundred shops. It had prosperous houses
with gardens, and many well-dressed notables. All around the city were
gardens, with roses, vineyards, and vegetables; outside it, in other gardens,
was a domed stone bath, with hot water of some value for cures.98 Some
parts of this description, notably the history of the castle and the repetition
of the number seven, seem conventional, but the overall impression is of
considerable commercial prosperity and a large population. The seventeen
districts recall the seventeen mosques restored by Jebrail ibn Jaja, and raise
the possibility that the town had always been so divided. If so, its size and
prosperity might have been continuous through the intervening centuries;
but, lacking sources, this must remain a speculation.
The contemporary Haji Kalfa is less explicit, merely recording that
Eskişehir was the seat of a kadi, had a small market, and contained the tomb
of Sheikh Edebali.99 European travelers are uninformative, most of them
merely recording passage by or through the town which seems in their time
to have offered no more attractions than it does now. Often, however, they
discuss the strategic importance of the site, and comment on its division
into two discrete settlements, one around the castle, the other, with the
market, near the baths.100
In most cases here studied, it has been possible to determine something
about the existence or condition of a given town or fortress in the time of
Osman, but necessarily far less about the countryside, rarely mentioned by
contemporary sources. For this, the endowment deed of Jebrail ibn Jaja pro-
vides some remarkable and exceptionally apposite information. The docu-
ment lists in detail the property which he left to endow his mosque; it gives
not only the names of the villages involved but a striking image of local
agriculture and trade.
The main property consisted of the village of Kara Gova, which was
bounded by the lands of villages called Eğri Özi, Alıncık, Göç Özi, and
another whose name has not been read; the village of Göç Özi which
stretched to Direkli, Saru Kavak, and the road to the city; and miscellaneous
property to be considered shortly.101 The donor specified that these villages
were given with everything they contained, namely: land, houses, shelters,
wells, streams, fruit trees, and anything cultivated or planted; as well as any-
thing usable such as meadows, towers, plains and hills, pasture, timber,
tools, and vehicles. Although the list contains elements which are no doubt
formulaic in such documents, it certainly indicates a flourishing agricul-
tural economy.102 The other specified items confirm the impression. They
include mills, a vegetable garden, and several pieces of land; a large house
with a portico and eight rooms; and two caravansarays. One of the cara-
vansarays was given with its shops and their contents, which included cloth,
and material of silk and wool.
The country was evidently well organized and exploited: mills ensured
irrigation in this dry region, which could then produce fruit, vegetables,
and other crops, as well as timber and pasture for animals. Property had
evidently been measured and delineated on a large scale. It is not surprising,
considering the strategic location of Sultanyuki, to learn that trade also was
flourishing. The caravansarays in the villages, presumably on main high-
ways, as well as that in the town, indicate the passage and presence of
numerous merchants, while the contents of the shops give a hint at the
goods exchanged. Wool and cloth were presumably local products, from the
sheep which abounded in the region; the silk may have been brought in
from outside, either from Byzantium or from one of the Seljuk cities where
it was woven.103 The caravansarays have not survived, nor others preserved
in the immediate region, but caravansarays of the thirteenth century at

Seyitgazi to the southwest, and between Kütahya and Afyon to the south-
east, stand on roads which lead to Eskişehir.104
This information is valuable enough in itself for giving an impression of
conditions somewhere in the large region around Sultanyuki. Yet the docu-
ment is not abstract, but gives specific names. When these are attached to a
region, the details become even more vivid and striking, and gain an
important implication. In fact, with the aid of a large-scale map, it is possi-
ble to identify the area in question quite precisely. The village name Eğri Özi
still survives in the form Eğriöz, the only place in the whole region which
bears such a name. This alone might seem a coincidence if the neighboring
village were not called Alınca, which plainly represents the Alıncık of the
document. Likewise, the name which the editor presented only in transcrip-
tion can be read without any major change as Söğütönü, which lies immedi-
ately south of Alınca. More speculatively, Direkli, “(the place) with columns”
might have been applied to the curious monument called Beş Kardeş, a
group of ancient stelae on a ridge west of Eğriöz, where the route to Söğüt
has long passed and Saru Kavak may be represented by the present Kavacık,
immediately east of Eğriöz. In any case, it appears that Karagova, whose name
does not survive, was approximately on the site of the modern Keskin.105
Villages of this region, as well as many others scattered through the prov-
ince of Sultan Önü, appear also in a list of local vakfs from the time of
Mehmed the Conqueror, compiled probably around 1470. Although it pro-
vides far less detail of individual sites, this document also gives an over-
whelming picture of agricultural prosperity, with large numbers of villages,
and extensive irrigation works.106 It would appear that little had changed in
the intervening two centuries and, in fact, the presence of thirteenth-century
village names on the modern map indicates a remarkable degree of conti-
nuity in the region since the Turkish occupation. The major break had come
before; the document of the fifteenth century, like the modern map, is strik-
ing in its lack of pre-Turkish toponyms. It would seem that the century of
chaos and conflict between the arrival of the Turks and the Seljuk acquisi-
tion of Dorylaeum had effectively destroyed earlier traditions and provoked
a major change of population.
The region thus identified lies immediately northwest and north of
Eskişehir at a distance of eight to fifteen kilometers, on the slopes of moun-
tains which separate the plain from the valley of the Sangarius, overlooking
the road from Söğüt. If, as it appears, the road passed through It Burnu in
Osman’s time (as it did in the nineteenth century), it passed through this
district, which is thus the immediate neighbor to the homeland of the

Ottomans.107 The countryside here is drab and empty, a treeless plain and
bare foothills now occupied by numerous villages. When cultivated, it is
productive of wheat, but suffers from a shortage of fuel.108 These open
steppes would seem ideal for nomads, and, if one were to find them any-
where near Söğüt in the time of Osman, this land would seem the obvious
place to look. Yet, on the contrary, they are conspicuous by their absence in
the late thirteenth century. The peaceful agricultural existence which the
will of Jebrail ibn Jaja implies testifies more than anything else to the great
achievements of the Seljuks of Rum in this district, given over to nomadism
when they took it from the Byzantines a century before. The country had
evidently been tamed, exploited for agriculture, and kept under central con-
trol; the setting for nomads is nowhere apparent. That does not mean, of
course, that nomads did not exist somewhere else in the district. They had
been there in the twelfth century, and were presumably settled on the land
by the Seljuks or driven out to pastures not under the control of the govern-
ment, and they may well have reappeared later in more chaotic days. The
notice of al-Umari about the lack of a large city and the presence of exten-
sive pastures may hint at a revival of nomadism (though his mention of
sprawling villages seems incompatible with that), but it seems not to have
been a major ecological factor in the region immediately adjacent to the
broad plain closest to the hills and valleys of Söğüt at the time when then
ancestors of the Ottomans were settling there. Sultanyuki or Eskişehir thus
appears at a center of trade and agriculture in the time of Ertuğrul when it
was still subject to the authority of the Seljuks/Mongols. Later, during the
reigns of Osman and Orhan, it was the capital of an independent principal-
ity which presumably still prospered, perhaps not as it had. The tradition,
which attributes the conquest of Eskişehir to Osman is plainly mistaken or a
fabrication. This tradition, which I generally follow here because it is by far
the most detailed, is based on sources of the late fifteenth century. The earliest
Turkish sources, however, whose narrative is generally too sketchy to serve as
a base, give a far different account. According to Ahmedi, who wrote at the
beginning of the fifteenth century, Sultanönü was conquered not by Osman
or even Orhan, but by Murat I together with Ankara, and thus in 1362 at the
earliest.109 In view of the known history of the city, this later date is certainly
to be preferred, and Sultanyuki/Eskişehir seen as independent (or ruled by
Germiyan) through the reigns of Osman and Orhan.
The list of vakfs from the reign of Mehmed the Conqueror provides an
indirect confirmation of a later dating. Most of the properties it lists were
donations of Mehmet II or his father Murat II, some go back to Beyazit I

and a few apparently are of Murat I. Only one purports to be older,
and that rouses suspicion by claiming to bear the seals of Osman and
Orhan. The province of Sultan Önü thus contained a great number of
properties donated or confirmed in possession by sultans of the fifteenth
and late fourteenth centuries, and none prior to the reign of Murat I
(1362–1389). It clearly indicates establishment of Ottoman rule by the
latter, at the earliest.110
The discussion of these few conquests of Osman has led a long way, but
may produce a history of the region, at least in outline. The district later
called Sultan Önü, with its main centers of Eskişehir, Karacahisar, and
İnönü, contains the broad plains which lie closest to Söğüt. It appears to
have remained under precarious Byzantine control in the twelfth century,
when the Comnenian emperors had to defend it against the encroachment of
Turcoman nomads. The fortresses of Kandilli and Karacahisar are monuments
of this age, prior to the great advance attempted by Manuel Comnenus,
whose refortification of Dorylaeum represented a major effort to establish
Byzantine control over this strategic region. His fortunes collapsed in the battle
of Myriokephalon in 1176, after which the Seljuks immediately occupied
Dorylaeum, thereafter Sultanyuki. They presumably took over the other
fortresses, and consolidated their power by constructing the defenses of
İnönü. Sultanyuki and its immediate region prospered under the Seljuks/
Mongols through the late thirteenth century; agriculture and trade spread,
nomadism was reduced.
In 1266 and 1272, the Seljuk Kaikhusraw III nominally ruled the area,
but the Mongols—the Ilkhanids whose empire stretched from eastern Iran
to Anatolia—were really in control. A powerful and fearsome empire stood
adjacent to the insignificant realm of Ertuğrul, whose son Osman thus
came to manhood in the shadow of the descendants of Genghis Khan.
The installation of Germiyan in the immediately adjacent region of
Kütahya in 1277 would have interposed a new power between Mongols
and Ottomans especially around 1300 when Germiyan was at least
temporarily in control of Ankara. A generation later, in 1326, Sultanyuki
appears as the center of a state, possibly an independent principality, or
possibly ruled by a cadet branch of the Germiyanids. The region, in any
case, was not Ottoman, and came under their control only in the reign of
Murat I. The curious inscription of İnönü seems to suggest that town
remained independent somewhat later. In all this, the tradition is seriously
defective or actively misleading.
1292: Köse Mihal of Harmankaya and Raids

Across the Sangarius (APZ cap. 10)
Osman’s frequent companion on campaigns was Köse Mihal, the Christian

lord of Harmankaya. On one occasion, Osman proposed a raid on Tarakçı
Yenicesi; Mihal responded with enthusiasm, suggesting a route via Sorkun,
Sarıkaya, and Beştaş, where the Sangarius could be crossed. He proposed a
further attack on the rich province of Mudurnu, and that Samsa Çavuş be
invited to join forces with them. The Ghazis followed this plan and arrived
at the tekke of Beştaş where the local sheikh told them they could cross the
river.111
On the other side, they met Samsa Çavuş, who had originally come to
Söğüt with Ertuğrul, but had moved to Mudurnu because of trouble with
the infidel ruler of İnegöl. The united warriors then attacked Tarakçı
Yenicesi and Göynük. From there, they moved on to Göl Kalanos before
returning to Harmankaya and Karacahisar, Mihal serving as guide through-
out. They took no captives, but came away with a great quantity of booty.
Such campaigns were intended to bring the local people into subjection.
Osman nevertheless maintained his good relations with the Christian lord
of Bilecik.112
This narrative is of considerable significance as an account of the preda-
tory attacks on their neighbors from which Osman and his followers may
have derived a substantial part of their revenues. One of the protagonists
may be real; the places certainly are, set in a wild landscape which suitably
illustrates the events and the society which they imply.
The name of Köse Mihal, one of the most important early followers of
Osman, suggests that he was, at least in origin, a Christian. Inscriptions and
documents of the fifteenth century and later name other Mihals who have
been identified, with varying degrees of plausibility, as his descendants. One
of them, Gazi Mihal, founded a mosque in Edirne in 1422 and was buried
next to it. His family and descendants were active in the conquest of Rumeli
and in the service of later sultans. Their connection with Köse Mihal, how-
ever, seems highly specu1ative.113
The case of Mihal Beg of Gölpazar is more germane. This worthy erected
a complex (to be considered next) including a caravansaray whose inscrip-
tion indicates the rather elevated titles he employed when the building was
founded in 1418.114 Documents of the late sixteenth century reveal that he,
there called Gazi Mihal Beg, controlled a large area north of the Sangarius
with many specified villages. At that time, the property was in the hands of
Boyali Mehmet Pasha, vezir and beylerbey of Aleppo, who died in 1593. His
grandfather had bought it from the heirs of the grandson of Mihal Beg, a
lapse of time which indicates identity of this man with the builder of the
caravansaray.115 The large territory and the titles which Mihal claimed show
that he was a man of considerable wealth and influence in the district.
Since this Gazi Mihal lived in the early fifteenth century, he is clearly not
Köse Mihal, but some connection seems inevitable, for he had the same
name and lived in the same area. Gölpazar is some twenty kilometers north-
west of Harmankaya (in a straight line which cannot be followed on the
ground), and many of the villages of the sixteenth-century document can be
identified. They show that Mkhal Beg controlled the Sangarius valley from a
point northeast of Söğüt and about fifteen kilometers from it, eastward to
Gömele (now called Mihalgazi), which is at the longitude of Eskişehir. His
lands stretched north to include Harmankaya and Sorgun. This is the home-
land of Köse Mihal, and immediately adjoins that of Ertuğrul and Osman.
The whole area was associated with the family of Mihal, evidently a large
and powerful one, for the vakf register of Sultanönü of 1472 identifies
Gömele as “Mihallerde” that is, “among (the lands of) the Mihals.”116
This close connection raises the obvious possibility that Mihal Beg was a
descendant of Köse Mihal, a supposition generally accepted, with the later
Mihal considered as the grandson of the earlier. Although this presumes
remarkably long generations, the line of descend seems plausible enough
until further reflection invites skepticism. Neither Mihal of Gölpazar not his
namesake of Edirne makes any reference to ancestors beyond the preceding
generation, an exceptionally curious circumstance if either was descended
from one of the most famous figures in their history.117 This means that the
existence of Köse Mihal is attested only in the tradition.
Parallels from early Roman history suggest a different solution, not
encouraging for the historicity of Köse Mihal. Roman historiography in its
formative stage was greatly influenced by patrons in the late Republic who
were fond of seeing their ancestors in a prominent role in the past, where
they could oblige the historian by filling the uncomfortable void before
written records. Many people and events were thus retrojected, only slightly
changed, to provide a coherent and continuous narrative from the earliest
days. Such a process seems indicated here. The powerful family of the
Mihalloğulları dominated the area around the Sangarius at the very moment
when Aşıkpaşazade was staying in Geyve (about thirty kilometers due north
of Gölpazar) and learning much of his history from Yahşı Fakih. They could
have made their influence felt on the tradition, even supplying material to
the historian or his informants. Alternatively, the Mihaloğulları of Edirne,

who were far more prominent in court circles, could have imposed a real or
imagined connection with Harmankaya on the tradition while it was being
organized in the capital. The shadowy figure thus created could have had no
greater glory than to have been a companion of Osman; his Christian origin
(and subsequent fortunate conversion to Islam) could have been merely
derived from his plainly Greek name. These, of course, are mere supposi-
tions, but no more unfounded than the existence of a historical Köse
Mihal.118
Whatever the historicity of Mihal, the scenes of his activity existed, and
contain remains of some interest. His home, Harmankaya, is represented by
the village of Harmanköy which stands at the northern end of an isolated
valley surrounded by mountains. Most spectacular among them is the sheer
rock face of the so-called Harmankaya, which rises several hundred feet
directly behind the village (Fig. 1.14). It towers so far above the level of the
surrounding hills that it is clearly visible from the tomb of Ertuğrul outside
Söğüt, giving rise to local stories about communication by signal between
Osman and Mihal. Although isolated and difficult of access, the valley lies
on routes which connect the Sangarius (and Söğüt) with the plain of
Gölpazar and the region of Taraklı and Göynük. The approach from the
south rises up a long and steep pass, suddenly offering a breathtaking view
Fig. 1.14 Harmankaya, home of Mihal

of the valley far below and the sheer cliff behind. To the northeast, an equally
steep climb leads through rough mountain country with hardly any flat or
open ground and only an occasional village until a watershed is crossed and
the country becomes softer and greener, gently sloping toward the marshy
plain of Gölpazar. The road to the northeast, which I have not followed, is
supposed to be equally steep and rough as far as the town of Yenipazar.
The valley of Harmankaya lies at an average altitude of 1000 meters and
is well-cultivated, much of it orchards of fruit and nut trees which flourish
in its cool climate. The villagers here (as in Gömele on the Sangarius) said
that they kept no livestock because the small area of cultivable land barely
sufficed to grow enough food for themselves. In earlier days, when agricul-
ture and communications were less advanced, the local economy would no
doubt have been marginal, supporting only a small population, cut off from
the rest of the world for two or three months during the winter.
Remains of a “Byzantine” fortification at the foot of the Harmankaya have
been reported and illustrated.119 Fragments of antiquity were manifest in
and around the village. In one place, numerous spoils seem to indicate the
presence of a Roman building, while carved stones and inscriptions of that
period were preserved in the village fountain, and at a tomb in fields outside
the village.120 The tomb, a simple burial in a walled enclosure, is tradition-
ally that of Köse Mihal. It is evidently of some antiquity, and incorporates
spoils of the Roman period, but there seems no way to establish its date. In any
case the remains show that the site was long inhabited, and was a sufficiently
prosperous town under the Romans to erect stone buildings and carve
inscriptions. A Byzantine period may be indicated by the remains of the
castle; the tomb provides equivocal evidence at best for the early Ottomans.
Osman and Köse Mihal followed a route to the region of Mudurnu which
led through Sorkun and Sarıkaya, crossing the Sangarius at Beştaş. They
returned via Harmankaya to Karacahisar whence, presumably, they had
started. One fixed point is Sorkun, whose identification with Çöte east of
Yenipazar and northeast of Harmankaya—on the route to Taraklı and
Göynük is determined by the sixteenth-century list of property of Gazi
Mihal.121 The other toponyms pose a more complicated problem. There is
indeed a prominent rock called Sarıkaya overlooking the Sangarius a few
kilometers west of Gömele, but the name is common. Beştaş has apparently
vanished, but its existence is attested in the fifteenth century, when the vakf
document of Sultanönü lists a farm held by the daughter of the dervish
of Beştaş.122 The dervish would have inhabited the tekke, or dervish
convent, of the tradition, and his existence leaves no doubt that such an
establishment existed somewhere on the Sangarius at a point where the

river could be crossed.123
The name, “Five Stones” suggests an ancient ruin, for which such a casual
designation is common.124 In fact, a ruin of the right kind was visible in a
suitable place at the end of the nineteenth century. Major von Diest in his
account of this part of the Sangarius valley, mentions the foundations of a
bridge on the right bank of the river just below the crag of Bozaniç Kaya
east of Gömele.125 It has been associated with an ancient road between
Eskişehir and Göynük, and indeed in von Diest’s day one of the few roads
through the ranges north and south of the Sangarius, from Eskişehir to
Yenipazar and beyond, crossed the river at this point.126 In this whole dis-
trict, the river is deep, its current strong. Because of the difficulty in fording,
a bridge is virtually a necessity. The traditional account suggests that no
bridge was standing, but even in ruin its piers could have been suitable for
supporting a wooden structure or for anchoring ropes by which a crossing
mechanism could be pulled. Ibn Battuta describes such a system, in which
passengers and goods were placed on a kind of raft which was pulled across
the river by men on the opposite side, on the Sangarius near Mekece.127
Although the tekke has gone, this site would have been the best place for
crossing the river in Osman’s time. His campaign would thus have advanced
from here along the Sangarius below Sarıkaya and through the pass which
the modern road follows to Hermankaya and beyond.
The central Sangarius valley has a wild and romantic character, with a
feeling of isolation from the rest of the world, for the river here provides no
easy means of communication (Fig. 1.15). This section is totally different
from the lower course of the river with its good route through broad plains;
that begins at the confluence with the Karasu and will be described next.
South of the Roman bridge near the confluence, the Sangarius passes for
more than twenty kilometers through a series of narrow gorges with precip-
itous side which makes its course impossible to follow and necessitate long
and difficult detours.128 The river enters these near Karaviran, a village
northwest of Söğüt and two hours from it (in von Diest’s time) by a moun-
tain path. From here, for seventy kilometers to the east, the valley is consist
ently about a kilometer wide, gradually rising from 180 to 230 meters
altitude. The rich silt enables many crops to be grown, with the aid of irriga-
tion from the river; von Diest described a luxuriant subtropical vegetation
with mulberry, fig, cherry, walnut, and plane trees and many vineyards. Yet
the land is limited, and now allows for no surplus; the locals, as at
Harmankaya, keep no livestock. The narrow valley gives the impression of a
Fig. 1.15 The Sangarius valley, west of Inhisar
long, isolated basin surrounded by mountains which rise steeply from its
sides. Many high detached hills of fantastic shape dominate the landscape,
most notable among them Bozaniç Kaya near Gömele (Fig. 1.16). Von Diest
climbed its steep flanks to discover a fortification of polygonal masonry
without mortar, but reported nothing medieval.129 At that time, the valley
was sparsely inhabited—the largest town, Inhisar, had a population of 900—
and contained numerous remains, inscriptions, and coins. But for Osman,
like his successors until modern times, the valley was a place to be crossed,
not otherwise of significance. To some extent, this seems surprising, since
the great difference in altitude between the valley and Söğüt would seem to
make it a suitable winter pasture for nomads; but the sources are silent on
this aspect.
The places which Osman and Mihal attacked are well-known towns of
some size located, like most of those here studied, in a rough broken coun-
try, yet on a major line of communication. TARAKLI, the westernmost, is
the Tarakçı Yenicesi of the tradition, an identity established by the similarity
of name and the appropriate location (Fig. 1.17). It appears in history in the
fourteenth century when Ibn Battuta described it as a large and fine town
called Yanija (i.e., Yenice). It then lay in the territories of Orhan and had a
governor with a body of troops from whom an escort was provided for the
travelers. Ibn Battuta and his party lodged in the hospice of the local akhis,
whose presence suggests that the place had become Muslim.130 The town, in
a basin surrounded by hills, lies beneath a small round hill suitable for the
Fig. 1.16 Bozaniçkaya on the Sangarius
Fig. 1.17 Taraklı, a goal of raids
fortification which Evliya Çelebi described as built by the tekfur of Bursa

and captured by Osman; it was ruined in his day.131 If this description is
more than conventional, it indicates a Byzantine fort; but no remains have
survived, nor is the ancient or medieval name known. The site was already
occupied in antiquity, for Kinneir in 1814 saw broken shafts and capitals of
pillars (which are no longer in evidence).132 The most prominent monument

now is a large mosque of the sixteenth century, which indicates an importance
continuing into Ottoman times.
The location of Taraklı is remote, the country rough. A passable road
leads from Geyve on the Sangarius, soon rising through ever higher hills,
with long views back over the river valley. The country rapidly becomes
more broken and enclosed, with forests and occasional stretches of farm-
land. Most of the route is empty and seemingly desolate, but cultivation far
up the slopes indicates the presence of numerous villages hidden in small
valleys and basins. For most of the distance from Geyve, the stark crag of
Karakaya is constantly visible, typifying a landscape which reminded
General von der Goltz of views of Herzegovina in the illustrated papers of
his day. He heard stories of bandits who infested the region, a danger no
doubt endemic to most periods.133 A similar road leads eastward to Göynük.
This, too, climbs and drops sharply as it passes through difficult broken
country with extensive forests on hills of highly eroded conglomerate. The
erosion has produced much good soil for the valley which the road follows.
Here, too, dramatic scenery, bad roads, and bandits occupied the attention
of earlier travelers.134
GÖYNÜK is a substantial and highly picturesque town with elegant tim-
bered houses lining the slopes above the confluence of two streams
(Fig. 1.18). When Ibn Battuta came, it was inhabited by Greeks, with only
one household of Muslims, that of Orhan’s governor. The town then had no
trees or vineyards but produced only saffron.135 Evliya Çelebi reported an
empty ruined castle of the tekfur of Bursa, conquered by Osman in 1312.136
This has not survived, but the account of Ibn Battuta shows that Göynük
was a Byzantine settlement, no doubt meriting defense by a castle. Earlier
remains indicate a history which goes back to the Romans, while substantial
Ottoman buildings and travelers’ accounts show that the place long retained
its importance. Among them the mosque and bath of Suleyman Pasha in
the center of the town were probably not built in 1331–1335 as claimed, for
the contemporary description of Ibn Battuta indicates that there was then
no Muslim population to use them. They are, however, monuments of an
early Ottoman period.137
Two roads lead east from Göynük: one, followed by most travelers who
constantly complain of its difficulty, to Nallıhan and Ankara, the other,
northeast to Mudurnu and Bolu. This passes through forests on steep hill-
sides and a region with small deep lakes before reaching a valley which
gradually becomes broader and more fertile as it opens out into the plain of
Fig. 1.18 Göynük, once an emirate
Mudurnu. Around midpoint, it passes through near the village of Genbemüz

which has an old mosque and bath attributed to Samsa Çavuş. Although the
buildings are not of such an age, their presence suggests a long tradition
identifying the homeland of this shadowy figure.138
In the winter of 1333, Ibn Battuta followed this route through steep slopes
and mountains and along a river which he crossed more than thirty times.
At this point, his guide demanded money, then decamped leaving the travelers
stranded since the route was obliterated by snow. They found a great quantity
of stones which indicated the road, perhaps the remains of ancient paving,
but could make no further progress. Finally, Ibn Battuta set off by himself
and providentially found a hospice whose sheikh rescued and lodged
the party. The following day they reached Mudurnu, where the hospice of
the akhis was full, but they did have the good fortune to meet a native of the
town who had made the pilgrimage of Mecca and spoke Arabic. He led them
to the bazaar where they could tie up their horses and buy supplies.139
MUDURNU, as described by Ibn Battuta, was evidently a prosperous
town with a market and at least one citizen rich enough to have made the
pilgrimage and, as it turned out, to act as a not very honest local money-
lender. It had succeeded the Byzantine Modrene, about which virtually
nothing is known except that it had a bishop, and was therefore a city,
apparently the most important place in these parts. It prospered then as
now from the fertile agricultural land and extensive forests around, and
from is location at the crossing of routes to Bolu, Ankara, and the plan of
Adapazar. Of these, only the route to Bolu is relatively easy, for the town
stands in a basin surrounded by mountains. Monuments of the late four-
teenth century—a large mosque and bath of Beyazit I—attest to its impor-
tance under the early Ottomans.
The town clusters at the foot of two hills (Fig. 1.19). One of them, steep,
elliptical, and detached from the surrounding ridges, contains the remains
of fortifications. These comprise a stretch of wall, with fragments of towers,
Fig. 1.19 Mudurnu, once a bishopric

all consistently built of mortared rubble slightly more regular on the face
than in the core. They are extensively bonded with wooden beams, but use
no brick. Such an indeterminate style is difficult to date; since it seems to
correspond with nothing Byzantine, the walls may be assigned to an early
Ottoman period. There is in any case no reason to doubt that the town was
in existence and worth raiding in the time of Osman.
The goals of this raid were not chosen at random, but all have in common
a location on a major highway. In the eighteenth and nineteenth centuries, it
formed part of the main route between Istanbul and Ankara and was fol-
lowed by numerous travelers. They were universally distressed by the diffi-
culty, for the road was by then in bad condition and the country it passed
exceedingly rough. The problems of Ibn Battuta show that the situation was
hardly different in the fourteenth century, but once, it seems, the road has
been paved. The stones of Ibn Battuta no doubt correspond with the traces
of Roman paving observed by Lieutenant Anton in 1893 in Tarakli and
between there and Geyve.140 In that case, the route had a long past; it would
plainly have been used in the time of Osman.
Location of these places on a highway has another significance, with
implications about Osman and his activities. The towns were centers of
trade, with travelers and caravans passing along the road. They were thus
the natural goals of a predatory raid, for not only might rich booty be avail-
able, but the rough country would enable the raiders to strike without
warning and to disappear with little danger of being followed. Osman and
Mihal did not choose those places by chance, but were raiding a strategic
area outside their domains and doubtless still inhabited by prosperous infi-
dels. The example of Göynük shows that Islam penetrated these hills only
gradually from the more populous areas to the east and west. Samsa Çavuş,
if he existed, would have been a local mountain chief like Gazi Mihal. It is
hardly an accident that travelers complain so frequently of bandits in this
country, for so must the ghazis have seemed to the neighbors they robbed.
In their return, after inflicting damage on Göynük and Taraklı the war-
riors descended to a place variously called Göl Flanoz or Göl Kalanos. This
has been plausibly identified with Gölpazar, known merely as Göl in the
sixteenth century. The second name has been explained as that of a local
Christian chief, corresponding to Kalanos, brother of the tekfur of Karaca
Hisar already mentioned. GÖL PAZAR is the largest town of the region
north of the Sangarius, a prosperous market and administrative center.
It lies in a broad and fertile plain, still somewhat marshy where the lake,
from which it derives its name, has been drained. In 1893, von Diest
described the lake as surrounded by swamps and the plain as almost

uncultivated presumably because it had an unhealthy situation; agriculture
and villages were in the neighboring hills and valleys. Such a circumstance
is common to many parts of the whole region, where plains now fertile and
densely populated were once marshy and desolate.
Whatever the state of its territory, there is no doubt that Göl Pazar has
long been of some importance because of its location on the main Roman
highway between Constantinople and the east, the so-called Pilgrim’s Road
whose course through here is clearly defined by milestones.141 Although not
mentioned after the late fourth century, and replaced in Ottoman times by
the route through Tarakli and Göynük, there is no doubt that this road was
in constant use in the early fifteenth century. Its importance then is attested
by the caravansaray of Mihal Beg completed in 1418 as part of a complex
which included a mosque142 (Fig. 1.20). Thus, the town was by then Muslim,
and a center of trade and transit. These buildings, the largest in the region,
would seem to indicate Göl Pazar as the headquarters of Mihal. Nothing is
known of the previous thousand years, but the town was probably suffi-
ciently prosperous in the century before Mihal to merit an attack.
This simple account of a raid had led a long way, illustrating the tradition
and putting it to a test. Köse Mihal may have been not a historical figure
but the reflection of a later namesake powerful in the region. Samsa Çavuş,
likewise, need not have existed, though he did become the subject of local
Fig. 1.20 Caravansaray of Mihal beg at Gölpazar

tradition. Both, however, are set in real places whose characteristics are
significant. First, they are remote, in a difficult mountain country not capable
of supporting a large population of men or beasts unless in conditions of
peace and good communications. Even the central Sangarius valley offers
limited possibilities, being functionally as isolated as the rest. Local people
might seek elsewhere for profit, and turn naturally to the towns situated on
the major route of trade which passed just to the north through Gölpazar or
through Taraklı and Göynük. Predation, as the sources clearly reveal, could
have played a significant role in the life of such chiefs as Osman and Köse
Mihal who only differed in that the former came from a place on a highway,
his companion from a more remote district. Predation, of course, is as com-
patible with a nomad as a settled existence, perhaps more so; yet the country
hardly offered scope for much nomadism. If the population were small,
transhumance between the Sangarius and the high basin of Harmankaya,
for example, might have supplemented the agricultural life of the rest, but
the present situation suggests that it might have been a luxury in a region
with so little arable land. In the chaos which accompanied the fall of
Byzantium, and no doubt always characterized a frontier region such as this,
existence was probably marginal, and raids a necessary supplement to the
normal means of livelihood. The places raided, on the other hand, were set-
tled and apparently prosperous from trade. In all this, the traditional
account makes sense in the landscape, whatever the value of its details.
“1288”: Marriages and Massacre: APZ 11–12
This episode begins with the marriage of Köse Mihal’s daughter to the lord
of Göl Flanoz. All the neighboring infidels and tekfurs, as well as Osman,
are invited. Osman arrives last, bringing (typically nomadic) presents of
rugs, kilims, and flocks of sheep. The tekfurs, amazed by his bravery, see him
as a potential threat, but they find no occasion to seize him. Ostensibly,
Osman maintains close friendship with the tekfur of Bilecik, but in fact they
are suspicious of each other. Nevertheless, Osman continues to entrust his
valuables to the castle of Bilecik.
The tekfur of Bilecik plans to marry the daughter of the tekfur of Yarhisar
and invites all the neighboring tekfurs as well as Osman. Warned by Mihal
to be on his guard, Osman suggests that the tekfur move the celebrations
from Bilecik, a narrow place, to Çakır Pınar, a suggestion the tekfur accepts.
On the night before the wedding, Osman’s men, hidden in the wagons that
were bringing his property, kill the guards, and seize the castle. Others,
dressed in women’s clothing, enter into the celebrations. Osman then rides
off, followed by the drunken tekfur, to Kaldırık, a valley near Bilecik, where
he decapitates the tekfur. In the morning, he captures the tekfur of Yarhisar,
with his bride and the wedding guests. Osman sends Turgut Alp to İnegöl
where the Turks kill its tekfur and his men, making captives of the women,
getting revenge on a man who had been responsible for the death of many
Muslims.
These activities represent a major advance from the hills of Söğüt to the
plains of İnegöl and Yenişehir, laying the basis for a serious threat to the
security and communications of Byzantine Bithynia. Bilecik was the key.
Having it in his own hands rather than those of the previously friendly but
now treacherous tekfur opened the way for expansion west to the plains or
north to the Sangarius. When APZ narrates Osman’s proposal to the tekfur
to move his celebrations from the confines of Bilecik to Çakır Pınar, he
shows a striking knowledge of local conditions, for the road leading west
from Bilecik almost immediately leaves the rugged mountains for more
open rolling hills with cultivated fields and scattered woods on the five
mile stretch to Çakır Pınar (which itself preserves no monuments of this
period) (Fig. 1.21). Kaldırık Dere, where the tekfur was killed, has not
been identified, but presumably lay between Çakır Pınar and Bilecik.
Further west, the rough country resumes until the plain of İnegöl is
reached, about ten kilometers from the town. In Evliya’s time, this road
passed through prosperous villages on mountains and slopes, but was
dangerous because of bandits.
YARHISAR was evidently an important place at this time. The modern
village sits amid high hills which start to rise soon after the road leaves
Yenişehir to the southeast (Fig. 1.22). Yarhisar is distinguished by its
mosque of Orhan, a roofed rectangular structure with a veranda overlook-
ing the village, its masonry very regular and its minaret described as one of
the finest of the period143 (Fig. 1.23). On the rocky butte above, the French
team found walls of mortared rubble and brick, with pottery of the thir-
teenth century, suggesting that this was a fortress of the Lascarids or their
immediate successors.144 Despite its apparent importance, Yarhisar seems
not to be mentioned by travelers. Instead, a major road led south from
Yenişehir with a stopping point at Akbıyık, some five miles southwest of
Yarhisar. This was the route used by Suleiman the Magnificent and his army on
their way to campaign in Iraq in 1534 and by Hans Dernschwam twenty-one
years later.145 Dernschwam complained about the bad road with constant
Fig. 1.21 Çakır Pınar
Fig. 1.22 Yarhisar

Fig. 1.23 Yarhisar: Orhan’s mosque
rises and descents through rocky mountains covered with scrub oak most of
the way to the plain of Ermeni Pazar, a description that would suit Yarhisar.
“1299”: Osman Proclaims his Independence: APZ 14–15
This momentous event supposedly took place not in the new conquests but
in Karaca Hisar, which was empty at the time of the conquest. Osman reset-
tled it, converted churches into mosques and established markets. The peo-
ple asked for a Friday mosque and a kadi (religious judge). They wanted the
pious Dursun Fakih but he told Osman that he would need permission
from the Seljuk Sultan. Osman rejected this, saying that he had conquered
the land with his own hand and had no need for any sultan. So Dursun
Fakih became kadi and led the prayers for Osman that symbolized inde
pendence. Osman established a tax on market transactions (according to
the story he was totally innocent of taxation) and distributed lands to his
followers in hereditary timars, land grants from which they would provide
fighting men when needed.
Osman’s Timars: APZ 16
Osman consolidated his conquests by distributing them among his follow-

ers to administer. The oldest conquests, Karacahisar and İnönü, went to his
son Orhan.
Others received Yar Hisar, İnegöl, and Bilecik. Osman himself chose
Yenişehir where he took up residence and had houses built for his followers.
This was a new town, the capital of a new realm.
Orhan now began to strike out in all directions together with his father.
They descended on Iznik and the Marmara region, bringing the infidels
there into submission. Several times, they attacked Köprü Hisar, but only
later conquered it. After each raid, they returned to Yenişehir.
YENIŞEHIR, “New City.” Osman’s new capital is now a dusty sprawling
town, at the edge of a large and fertile plain. Dernschwam remarked on a
curious rectangular tower of three stories with vaults all round and the large
caravansaray that reflected the volume of trade that passed through.146 For
Evliya in the next century, it was a thriving town with 1600 houses, seven
mosques, a soup kitchen, a covered market, and 150 dependent villages.147
In fact, Yenişehir preserves a bath traditionally attributed to Osman and a
mosque, a medrese, a monumental tomb, and a dervish lodge from the time
of Orhan.148
Yenişehir sits in the northern part of a plain that extends east and west
and is only separated from the plain of İnegöl to the south by a series of dry
ridges. It is a strategic location, with roads leading to Nicaea over a steep
mountain pass, to Lefke and the Sangarius along the valley of the Göksu, to
Bilecik where the plain yields to rugged mountains, and to Bursa over ridges
that separate the two large plains. For many, the plain of Yenişehir sup-
ported extensive cultivation of grain: Dernschwam found it rich but poorly
cultivated because of the laziness of the Turks; for MacFarlane it was “mag-
nificent expanse of the finest of corn lands.” To some extent, this was a
deceptive image. Dernschwam gives a hint by noting a swamp or lake he
saw as he traveled up into the hills. Pococke reported a great lake extending
over the plain, becoming a morass overgrown with reeds in the summer.149
For Texier, Yenişehir was built on the shore of a swampy lake, and Mordtmann
reported that the swampy plain made the neighborhood unhealthy.
MacFarlane noted that French maps showed a lake where there was no lake
at all. In modern times, the town was famed for a rich clotted cream (kaymak)
made from the milk of water buffalo, beasts who thrive in soggy or marshy
landscapes.150 This phenomenon of disappearing lakes is typical of the
region: depressions between ridges tend to be poorly drained, so that water
accumulates, producing temporary swamps or lakes.151
Lake or not, the communications of the site made Yenişehir a suitable
location for the headquarters of a leader whose ambitions stretched toward
Byzantium.
KÖPRÜ HISAR, some five miles east of Yenişehir at the edge of the plain
defends a river crossing and the road that leads over the mountains to the
Sangarius valley through sharply rising terrain that passes Balcık Hisar, in a
broad yayla high above Lefke. Köprü Hisar has the bridge that gives it its
name, an old hammam, and remains of fortification walls of brick and
mortared rubble, which would have enabled it to resist Osman’s and
Orhan’s attacks.152
“1302”: Defeating a Coalition: APZ 17
Evidently realizing that they were under serious threat from the Turks
established at Yenişehir, the commanders of the Bursa region organized a
joint campaign. The tekfurs of Adranos, Bidnos, Kestel, and Kite advanced
from the plain of Bursa through the hill country that led toward Yenişehir.
Osman gathered his forces at Koyun Hisar and moved on to Dimboz where
he gained a decisive victory. Although his nephew Aydoğdu fell, the tekfurs
were devastated: Kestel was killed, Bursa retired behind his own powerful
walls, and Adranos fled the scene as did Kite whom Osman pursued far to
the west, to Ulubad. Osman threatened the local tekfur with a devastating
attack unless he turned over the ruler of Kite. Ulubad agreed on terms that
neither Osman nor his descendants cross over the great bridge over the
Macestus river, an agreement that Aşıkpaşazade recounts was still observed
in his own time (but the Ottomans did cross the river in boats). As a result,
Kite’s tekfur was killed and his fortress taken by Osman. This victory, called
the campaign of Dimboz, brought Ottoman control far to the west, posing a
potential threat to Bursa, for Kite lies some 15km west of the city, poten-
tially controlling the roads that lead to the Dardanelles and interior Mysia.
Once again, these are all real places. ADRANOS is the farthest from the
main scene of activity. To reach Bursa, its forces had to advance through
rolling hills with cultivated fields, then pine woods and through a spectacular
pass—a break in the massive granite rocks of Kapulu Kaya—before a
rough descent to the Bursa plain. The town itself, on an isolated hill by the
river Rhyndacus, stood some two kilometers from the ancient site of
Hadrianoi, whose name it preserves. Its castle, of about 70 × 100 meters with
twelve towers and circuit walls over eight meters high, is now in ruins, as it was
when Hamilton saw it in 1836. Its brick and rubble construction, as well as
pottery found on the site, are consistent with a Lascarid date. The rough
mountain road to Bursa was well protected by two castles—near Kesterlek
whose brick and stone masonry resembled Ulubad, and at Kermasti where
Texier reported a Byzantine fort.153
BITNOS has not been located with certainty, though one document of
the sixteenth century names a Bednos Alani in the region of İnegöl.154
KESTEL, on the other hand, is well known. On the edge of the plain
about halfway between Dimboz and Bursa, its partially preserved walls
employ mortared rubble with brick courses in a style appropriate to the
Lascarids. It was already in ruins when Evliya Çelebi visited in 1672. He
curiously writes that Orhan conquered it in 753 (1352). It was a more
important site than at first appears, for MacFarlane pointed out that the
ridge where Kestel sits commanded the pass for the only road that led into
the interior of Asia Minor. It was “crowned by the picturesque ruins of a
castle, a work of the Lower Empire.”155
The fortress of KITE, in the plain sixteen kilometers west of Bursa, con-
sists of a well-preserved pentagonal structure some 130m on its longest
side.156 Hasluck described it as having triangular, pentagonal, and U-shaped
towers and only one gate. Its homogenous construction of rubble with
irregular bands of brick would suit the Lascarid period; pottery points to
the early fourteenth century. It has been identified (perhaps wrongly) by
similarity of name with Katoikia.157
The location of ULUBAT made it a suitable goal for Osman, though this
time he threatened rather than conquered (Fig. 1.24). The fortress is second
in size (475 × 150m) only to Bursa in this region, and controls the strategic
crossing of the Rhyndacus river (Koca dere) where it issues from Lake
Apolyont.158 Its stone Roman bridge avoids a detour around the lake, a
three-day march. The relatively well-preserved walls are the product of the
Byzantine emperor John Comnenus (1118–1143), built to consolidate his
control of the region against the Turkish attacks that had menaced it. Towers
and walls share a common masonry of courses of mortared rubble alternat-
ing with bands of brick, typical of the twelfth century. This fortress, key to
Fig. 1.24 The walls of Ulubat/Lopadion
the fertile Mysian plains as well as the major road into the interior, only
became Ottoman in the reign of Orhan.
This campaign, then, makes sense. It involves places that existed in
Osman’s time and reveals a plausible Byzantine strategy to maintain control
of the plain of Bursa and of the roads that led to the coast as well as the
interior—and of Osman’s desire to consolidate his realm based on Yenişehir
and to open the possibility of expansion westward into the large and fertile
plain of Bursa.
Bursa Blockaded: APZ 18
The victory of Dimboz posed a threat to Bursa that became material in the
next episode. Osman, realizing that the powerfully fortified city could not
be taken by storm (and it was unlikely that the Turks possessed sophisti-
cated siege equipment), ordered the construction of two blockading forts,
one by the hot springs west of the city, the other on the opposite side.159 The
purpose of these was cut Bursa off from its countryside, as well as from
sources of reinforcement. They made it impossible for the infidel even to
stick a finger outside the castle walls. But it would be many years before the
Turks could take Bursa.
“1304”: Conquest of Malagina: APZ 20
This campaign took Osman and his followers through a spectacularly scenic
region, the heavily fortified valley of the Sangarius. Osman, encouraged by
his constant victories, decided on further adventure. After receiving Mihal’s
conversion to Islam, he led his forces to Leblebicihisar, whose tekfur submit-
ted without a fight. So did the tekfurs of Lefke and Çadırlı. They turned their
lands over to Osman who entrusted a small fort near Lefke at the mouth of
the valley of the Yenişehir river to his senior commander Samsa Çavuş.
Osman then moved on to accept the surrender of the tekfur of Mekece who
joined him in attacking Akhisar where, after hard fighting, its tekfur took
refuge in Kara Cebiş Hisar, a fort high above the Sangarius. Next, the tekfur
of Geyve decamped for Koru Dere where he was captured and brought to
Osman along with much loot. Finally, the ghazis took Tekur Pınar. Osman
stayed more than a month in the region, distributing the conquered lands to
his followers and ensuring peace and security for the population.
Aşıkpaşazade claimed that the land remained unchanged until his own time.
Although a couple of identifications remain in doubt, this is a coherent
account of an attack on the strategic central Sangarius valley, moving from
south to north. The first fortress, though, poses a problem: the only
LEBLEBICIHISAR known was a ruined castle in the district of Göl Pazar,
cited in a document of 1607.160 At first sight, this is far off any rational route
that Osman might take from Yenişehir, though not impossible if he were
starting out from Söğüt or Bilecik. Proceeding north from either of those
through rough mountain country would take him to the western part of the
lands of Gölpazar which, in the time of the document, stretched as far as
the Sangarius. Leblebicihisar, then, may be sought in the valley upstream
from Lefke.
LEFKE, which preserves the name of Byzantine Leukai, was a prosperous
town with 600 houses, five mosques and a small square castle, ruined when
Evliya Çelebi passed through.161 The fort, which has left no trace, presum-
ably stood on the broad hill that rises high above the Sangarius and contains
a street of elegant Ottoman houses. The place was famous for its quinces—
as the valley still is. The location is strategic, for it commands roads along
the Sangarius, to Yenişehir via the valley of the Göksu, and over a pass
to Nicaea.
ÇADIRLI has not been located, but an identification can be suggested.
General von der Goltz recorded what he called the remains of a Roman castle
then known as Eski Kale, on the right bank of the Sangarius about eight
kilometers downstream from Lefke.162 The location close to Lefke and

protecting the approach to that city’s small plain from the north makes it a
candidate for the Çadırlı of the text.
The small fort at the mouth of the valley of the Yenişehir Su (now called
Göksu), which came to be named for Samsa Çavuş has been plausibly iden-
tified with the fortress variously called Iki Kule or KULELER, about five
kilometers upstream on a ridge over a wide spot in the valley of the Göksu
where there is a bridge (Fig. 1.25). Its towers feature a distinctive masonry
that finds its counterpart in thirteenth century walls of Nicaea, a date that
conforms with the pottery found there.163
Returning to the Sangarius, MEKECE was protected by a small (30 × 70m)
poorly preserved castle about two kilometers west of the modern village.164
Von der Goltz reported it as a Byzantine fort rebuilt by the Turks; pottery
is of the early fourteenth century. This fort, which controlled passage
along the left bank of the Sangarius, offers a first view of one of the most
distinctive and important features of the region, the broad plain that
stretches some twenty-five kilometers along both banks of the river, from
Mekece to Lefke.
This plain, known as Malagina, played a major role in the military history
of Byzantium. It was the place where armies were mustered on their way
Fig. 1.25 The fort of Iki Kule

east, the seat of the imperial stables and by the thirteenth- century
headquarters of a province of its own. It is the largest open space in this
stretch of the Sangarius, for the river otherwise passes through a series of
gorges from Harmankaya to the plain of Adapazar, east of Lake Sapanca.
The fertile plain at Lefke is an exception, but on a small scale. Such a
strategic area called for serious defenses; they centered on the powerfully
built castle of Metabole, on a rise that reaches an elevation of 600m providing
superb views over the Malagina plain and the mountains beyond165
(Fig. 1.26). The steep hill, combined with a ridge that protects and conceals
it from below, make it difficult to access and easy to defend. Its best-
preserved wall, reinforced by triangular bastions, is faced with neatly
arranged limestone spoils—mostly column drums and tombstones—that
indicate a date in the Dark Ages and give the castle its suitable Turkish
name, the White Castle of the Sangarius (Fig. 1.27). Here, then, is the
Akhisar that only fell to the forces of Osman after a difficult struggle. The
tekfur of Mekece who joined him probably helped in finding a suitable
approach to the castle.
The tekfur of Akhisar took refuge in KARA ÇEBIŞ HISAR, high above
the Sangarius. The name does not survive in the region, but the remains of
Çoban Kale, which von der Goltz noted in the scenic gorge between Geyve
and Adapazar, above the traces of a bridge, may represent the site; it was
certainly a remote and defensible one. For Evliya Çelebi, it was a small,
Fig. 1.26 Malagina, overlooking the Sangarius

Fig. 1.27 Walls of Malagina/Akhisar
ruined castle on a steep rock on the banks of the Sangarius, three hours
from Geyve, where the road was so narrow that passersby could be forced to
pay a toll. The heavily overgrown remains consist largely of one tower
(much has succumbed to road building) with some brick bands amid its
fieldstone facing. Comparison with Akhisar has suggested a date in the late
twelfth century.166
With its main bastion captured, the rest of Malagina soon fell. Next to be
abandoned by is commander was GEYVE, the Byzantine Kabaia, mentioned
only in 1275 as a place of exile.167 Its suburb, now called Alifuatpaşa, is
graced by a stone Ottoman bridge (Fig. 1.28) above which rises an isolated
hill, seemingly ideal for a fortress. The local historian Namık Cihan, how-
ever, explained that there was nothing on top of the hill. Nevertheless, the
place did have a “very small castle” which the indefatigable Evliya Çelebi
saw in 1648.168 Geyve in his time was a small prosperous town which had
been much larger before being devastated by a flood a few years earlier.
Then as now it prospered from its rich fruit production. Koru Dere where
Geyve’s tekfur fled, has not been located, but a document allows TEKUR
PINAR to be identified with Umurbey, some five kilometers southwest of
Geyve at the edge of the plain.169 It has not been investigated. With that,
Osman’s conquest of Malagina, which opened access to Nicaea. Nicomedia
and the lower Sangarius, was complete.170
Fig. 1.28 The Sangarius at Geyve
Attacks to the North; First Threats to Nicaea

and Nicomedia; APZ 22
Osman now sent fighters in two directions: Orhan, leading a campaign of

his own for the first time, struck toward Kara Çepiş, Ab suyu, and Kara
Tekin, while Konur Alp made his base at Beşköprü before advancing
through the forests toward Ak Yazı. As a result, Osman secured this frontier.
Meanwhile, Akça Koca headed for Izmit. After taking Kara Çepiş by a strat-
agem, Orhan moved on Ab Suyu which surrendered. He left the local popu-
lation in place under peace and security. Next to be attacked was Kara Tekin,
but Orhan announced that his real goal was Iznik. Orhan called on the tek-
fur to surrender and stormed the fort when he refused, capturing and killing
the tekfur and returning with much loot to Osman in Yenişehir. The effect of
this campaign was to turn Kara Tekin into a fortress for blockading Iznik
and a base for devastating its lands and making life very uncomfortable for
its inhabitants.
AB SUYU, described as a small fortress below Kara Cebiş, is perhaps to
be identified with Adliye, a fort whose thick walls are faced with mortared
rubble and brick in a technique appropriate to the thirteenth century.171 It

stands near the exit of the long gorge from which the Sangarius flows into
the broad plain east of Lake Sapanca. If this were captured, the ghazis
would have access to an environment far different from the river valley
they had been overrunning. This would have allowed Akça Koca to penetrate
to the bridge of Justinian (BEŞKÖPRÜ) over the Sangarius (Fig. 1.29), a
monumental stone structure still intact, with an apsidal structure at its
east end that presumably represents the small fort or tower (burguncuk) -
another accurate local touch—where he paused before attacking
places in the great forest that stretched along the hills and mountains
far to the east172 (Fig. 1.30). Occupation of the lake shore near the
bridge would have opened the main route to Izmit. Likewise, capture of
fortresses on the Sangarius meant that Konur Alp could move east across
the Akova plain to AKYAZI whose fortress, at the foot of the mountains a
couple of miles south of the modern town, has been mentioned but not
described.173
KARA TEKIN, now called Karadin, has a special location on the route
through the pass that rises steeply from the Sangarius valley before begin-
ning its long and gradual descent toward Iznik.174 For much of its course,
the route is confined by steep hills on either side, but at Karadin opens out
into a small plain (Fig. 1.31). The settlement, built on a prehistoric mound,
was well fortified with a wall of about 180 × 150 meters; its masonry of
Fig. 1.29 Justinian’s bridge over the Sangarius

Fig. 1.30 Tower-like structure on the bridge
Fig. 1.31 Karadin/Trikokkia
mortared rubble and brick suits the twelfth- and thirteenth-century pottery
found there. Texier in 1833 saw walls with towers, but by von der Goltz’s time,
there were only a few ruins. In any case, this was the fortress that commanded
the route to Iznik, whose fertile fields stretch east from the city, only thirteen
kilometers away, vulnerable to any blockade from this direction.
“1326”: Bursa Surrenders: APZ 23
When news arrived that the infidel defenders of Bursa were suffering from
hunger and only wanted a pretext to surrender, Osman called on Orhan and
put him in charge of a campaign whose first step was directed against
Adranos. This was a war of revenge for the father of the tekfur had been
responsible for the death of Osman’s nephew Bay Hoca in the battle of
Dimboz in “1302.”175 When the tekfur heard that the Turks were coming, he
fled to the nearby mountains (the Elete Dağ of the sources has not been
identified), where he perished from falling off a cliff. The Turks burned the
fortress but took the local population who had surrendered under their pro-
tection. Orhan then returned to Bursa where he took up his headquarters at
Pınarbaşı on the south side of the great fortress, and sent Mihal to negotiate
terms. The tekfur paid 30,000 gold florins for a protected departure from the
city. The Turks escorted him to the port of Gemlik; the wealth he had accu-
mulated in Bursa was distributed among the ghazis who thereby became
very rich. Osman himself did not participate in the capture of Bursa because
he had a problem with his leg, but really because he wanted Orhan to win
glory while he, Osman, was still alive.176
Adranos has been met before, in the coalition of the tekfurs of the Bursa
region. Evidently it had survived unconquered for the subsequent twenty
years, perhaps because it was far from the action of that period. Its fate was
linked with that of Bursa with which it was directly connected, but how it
was supplied and what its relation had been with the larger city can only be
imagined. In any case, it seems not to have been in a position to relieve the
blockade that brought Bursa to starvation.
The tekfur left via the port of GEMLIK, an ancient foundation that preserves
Byzantine rebuildings of its walls, some of them with alternating bricks and
stones in their masonry, typical of the twelfth or thirteenth century.177
Evliya mentions the fine castle, a product of the builders of the walls of
Iznik, that stood on a high hill by the shore, and could, he thought, easily be
restored.178
The narrative includes an edifying discussion with a certain Saroz, described
as the vezir of the tekfur. He explained to Orhan the main reasons for the
surrender, including the abstract—that the Turks power was growing from
day to day while theirs was declining. Finally, APZ asks whether Osman
were still alive at the time of this conquest (for he had not appeared on the
scene of such an important victory): the answer he gives is that Osman was
suffering in his leg so could not be present. This question will be discussed
next.179 In any case, the Turks had achieved one of their major goals,
conquering the first of the three great cities of Bithynia.
Expeditions East and West: APZ 25–27
Konur Alp was sent toward the woods and plains of eastern Bithynia—to
Ak Yazı, the Konurapa region, Bolu and Mudurnu, whence he returned to
Kara Cebiş and Ab Suyu before setting out again. Akça Koca, meanwhile,
was entrusted with an attack on Kandıra and Ermene as prelude to a
supremely bold mission, to strike at the environs of Istanbul through the
Kocaeli peninsula. After taking Kandıra, where he stationed some men, he
proceeded against Samandıra. Here, he encountered much resistance until
he surprised the local tekfur who had gathered with his men for a funeral
outside the castle walls. The tekfur was captured and Samandıra became
Akça Koca’s base for attacking Aydos. Here he encountered serious resist
ance from the local tekfur as well from the “tekfur of Istanbul” determined
to keep the Turks from Aydos. Akça Koca brought the captured tekfur of
Samandıra to the castle of Aydos, requesting ransom from the defenders
and inciting them to surrender. When they refused, he tried to ransom the
tekfur to Istanbul, but those infidels also refused to pay, so he finally sold
him to the tekfur of Izmit after much victorious fighting that finally brought
peace and security to the villages around Aydos. The fortress continued to
hold out until the tekfur’s daughter betrayed it as the result of a dream.
The expedition across the Kocaeli peninsula (which the Byzantines called
Mesothynia) was one of the boldest the Turks had undertaken. The first
goal, KANDIRA, is a major market town some fifty kilometers north of
Izmit across a hilly country. Until the late nineteenth century, this region
was believed to be covered by the dense forest that the Turks called the “sea
of trees.” Von der Goltz, however, by making an excursion to an area west of
Kandıra, found that it was a hilly country with many villages and settled
cultivation.180 Travelers seem not to have visited Kandıra.181 The town con-
tains a simple rectangular mosque named for Orhan and, on its outskirts, a
small wooden mosque and attached tomb both supposedly of the region’s
conqueror Akça Koca (Fig. 1.32). In any case, the region of Kandira con-
tains a group of distinctive small rectangular wooden mosques (built with-
out nails) which form a group attributed to Orhan and thus attesting the
earliest Turkish settlement in the area.182 ERMENE (presumably the same
as the Ermeni Pazar of APZ 30), mentioned in association with Kandira,
Fig. 1.32 Monument to Akçakoca in Kandıra
poses a problem. The obvious identification is with Armaşa (now called

Akmeşe), ten kilometers north of Lake Sapanca, the seat of an important
Armenian monastery and pilgrimage goal. Unfortunately, the monastery
was only founded in 1610 and nothing is known of the earlier history of the
site.183 In any case, Kandıra was only a stopping point on a much more
adventurous route that would take the Turks through steep hills and valleys
the whole length of the peninsula to Samandıra, only some twenty kilome-
ters from the Bosporus. By following such a route, they could hope for an
element of surprise, at the same time avoiding the main highway from Izmit
to Istanbul along the south coast, which was heavily fortified.
SAMANDIRA, now swallowed by the suburbs of Istanbul, was the
Byzantine Damatrys, site of a palace which Byzantine emperors through the
Middle Ages frequented for recreation and hunting; it offered a convenient
location in the country but only a short ride from the capital.184 It is last
mentioned in 1296, when Andronicus II camped out here for the summer
when Istanbul had been devastated by an earthquake. No castle is recorded
on the site, but the palace was massive enough, perhaps, to give the impres-
sion of a fortress.
AYDOS, on the other hand, is well known and has even the subject of
excavation.185 This formidable castle stands on highest peak of the region,
now in the midst of a small pine wood with the suburbs of Istanbul lapping
at its feet. The oval castle which occupies the hilltop, consists of a double wall
with 13 U-shaped towers and three gates. It is faced with rough fieldstones
set in mortar, except for a tower near the main gate which makes decorative
use of brick. The excavators date the structure of the twelfth/thirteenth
centuries, with additions in the thirteenth; its most active period was the
thirteenth/fourteenth centuries, finds from the Ottoman period were lacking,
suggesting that the place was abandoned not long after its conquest.
Conquest of Izmit: APZ 30
Immediately after succeeding to supreme power Orhan distributed fiefs to

his family, then organized a major campaign against Izmit. The troops gath-
ered at Bursa, marched to Yenişehir and Geyve, and found Orhan’s son
Suleyman at Apsuyu. The gazis who had been in Aydos joined them from
their base on the shore of Lake Sapanca, whence all proceeded by the
ancient highroad to Izmit. The ruler of that city was a woman, Yalakonda,
related to the tekfur of Istanbul. She was also the ruler of Yalova where she
had a castle in the valley. Her brother Kalyon controlled the castle on the
ridge above, which the Turks of APZ’s day called Koyun Hisar. After her
brother was killed, Yalakonda surrendered Izmit on condition for safe con-
duct for herself and her followers, with their possessions. With the terms
agreed, the defenders embarked in ships, presumably to Istanbul. Orhan
entered the city, as did the warriors from Aydos who henceforth were sta-
tioned in Izmit. Orhan converted churches to mosques and made one of
them into a medrese; the Greek city became Turkish. Orhan distributed
timars: Kara Mürsel took charge of the shore with troops to make sure no
harm came by sea from Istanbul; others got Yalova, Ermen Pazar, and
Kandıra. This concluded the conquest of Kocaeli, Ak Yazı, and Konurapa.
The route of the joint expedition to Izmit makes sense with its rendez-
vous at Apsuyu on the Sangarius, for one of the main routes across Anatolia
runs through Izmit, and south of Lake Sapanca on its way east. The ghazis
would have joined it (as many travelers did) near the mouth of the Sangarius
gorge. When Orhan achieved his goal, conquest of the last of the great
Bithynian cities, he naturally took over the coast of the Gulf, including the
strategic site port of Yalova, the “gateway to Asia” of the Byzantines, termi-
nus of a natural route that led to Nicaea and the interior of Asia Minor.
Evliya describes the fortress of YALOVA as built by “the tekfur” and con-
quered by Osman (no doubt a lapsus for Orhan), who had such trouble
taking it that he had it demolished, though its foundations were still visible.186
The town itself had seven hundred houses and seven mosques, but its
location on the coast was unhealthy, subject to malaria. The castle in the
valley, at Yalakabad, had been conquered by Orhan and Kara Mürsel from
the infidels was a place where locals kept their sheep in the winter, and a
stop for bandits and merchants. This is presumably Yalova, while the
castle on the ridge would be Çoban Kale, the Byzantine Xerigordos which
occupies an exceptionally strategic location above the Yalakdere valley,
where the coastal plain meets the hills of the interior.187 The road, which
rises gradually from the coast, runs directly beneath the fortress, for the river
here passes through a gorge. Up to this point the landscape is subtropical,
with vines, olives, fields, and fruit trees in the valley and stretching up the
slopes. Beyond it, the country becomes rugged, with thick maquis and
less vegetation. The land rises and becomes wilder toward the south
where the road passes through a tangle of high hills that separate the
Gulf from the lake and plain of Iznik. Çoban Kale has been surveyed: it is
an ovoid structure of 180 × 120 meters with seven semicircular towers
that could be identified and probably another five. Its masonry, of flat stones
in rough courses, is not diagnostic but would suit the textual evidence
that suggests it was already standing at the time of the First Crusade and
rebuilt in the twelfth century. Pottery found on the site was of the thirteenth
century. As for the related castle “on the ridge above,” it has not been
identified; it can hardly be the Koyun Hisar already met on the route
from Yenişehir to Bursa, but evidently was another place with the same
name, “Sheep Castle,” that would suit Evliya’s description. It might possibly
have been the Kale Tepe discussed in the next section.
Izmit is well known, with plenty of monuments from the time of
Orhan, including, it seems, rebuilding of parts of the fortifications, where
presumably the men from Aydos would have been stationed. Excavation
of that site confirms this narrative, for there was virtually nothing found
from the Ottoman period, as could be expected if the place was evacu-
ated or abandoned soon after its conquest.
“1331”: Conquest and Settlement of Iznik: APZ 32–33
The blockading fort at Kara Tekin was doing its job so well that the people
of Iznik were completely cut off from their normal food supplies, as the
Turks occupied all the land around and distributed it as timars. They
couldn’t even go out onto the lake for fishing. Eventually, Iznik surrendered
on terms: those who wished to leave could leave and those who wanted to
stay could stay. The tekfur left by the Istanbul (north) gate, but most of the
people stayed behind. Orhan entered by the Yenişehir (south) gate, and met
the infidels who acted as if their ruler had died and they were receiving his
son. Among them were many women whose husbands had died in the
fighting or from starvation. Orhan gave them to his followers to marry and
settled them in houses in the city (Figs. 1.33, 1.34).
Osman rapidly imposed an Islamic image on Iznik: he converted the
main church into a mosque, a monastery became a medrese and he built a
soup kitchen by the Yenişehir gate. For a time Iznik became his capital.
The Anonymous adds some details.188 When the Turks saw that Iznik
could not be taken because it had water on all four sides, they built a block-
ading fort on the mountain wall toward Yenişehir and garrisoned it with
forty troops under the command of the brave and strong Daz Ali. The fort is
called Taz Ali Hisar; it has a high rock above it from which springs a source
of cool water. The infidels of Iznik, recognizing their desperate situation,
managed to send a message to the tekfur of Istanbul, for in those days he
still ruled Iznik. The tekfur thereupon sent a fleet that landed at Yalakova,
intending to march on Iznik and surprise the Turks. The surprise was on
them, however, for the Turks had a spy who reported the landing back to
Fig. 1.33 Looking down on Nicaea

Fig. 1.34 Lascarid tower of Nicaea
the forces at Iznik, who marched to the coast and destroyed the attackers.
When news of this reached Istanbul, the tekfur was in despair, and the
defenders of Iznik soon surrendered.
Kara Tekin, long the base for blockading Iznik, was not the only such fort.
Daz Ali is also a real place, located at the village of Dirazali, on the heights
four kilometers south of Iznik, overlooking the plain and the city. This is
actually a strategic location, guarding the rough mountain road from Iznik
to Köprühisar. The fort, still standing in the time of the Anonymous, was in
ruins by the early sixteenth century.189
Çoban Kale, discussed above, controlled the main route from the sea,
while a small fortress, Kale Tepe, high above the south shore of the lake at an
elevation of 835 meters overlooked the plains of Iznik and Yenişehir. Its sur-
viving two small towers exhibit a masonry comparable to the Lascarid. On
the lake shore, eight kilometers southwest of Iznik, the rocky promontory of
Karacakaya was reinforced by a medieval wall; pottery of the thirteenth
century was found there190 (Fig. 1.35). It thus appears that the city was
indeed blockaded from all sides and that starvation forced its surrender.
The buildings of Orhan survive, most notable among them the mosque
converted from the church known as St. Sophia at the exact center of the
city. The role of Iznik as capital, however, is not attested elsewhere; in any
case, the next chapter states that Orhan entrusted the city to his son
Suleyman.
Rounding off the Conquest of Bithynia: APZ 34
After settling Iznik, Osman sent Suleyman to Tarakçı Yenicesi whose people
had heard of Orhan’s justice. Since the Ottomans brought justice wherever
they went, their good reputation spread to places still unconquered. As a
result, Yenice, Göynük and Mudurnu all willingly surrendered on terms.
Fig. 1.35 Karacakaya, a blockading fort

The Turks brought security and prosperity; many of the country people
converted to Islam. The conquest of Bithynia was complete.
These places have been met more than once before. On the first occasion,
narrated in APZ chapter 10 and set in “1288,” Osman, together with his
Christian ally Mihal, attacked Göynük, Tarakçı Yenicesi, and Göl Flanoz.
This seems to have been a raiding expedition that brought much loot, though
APZ unconvincingly adds that they did this in order to bring the local people
under their control. In chapter 25, set after the conquest of Bursa, Konur Alp
is sent out to Ak Yazı, Konurapa, Bolu and Mudurnu, taking the Turks a
good deal farther east. This also appears to have been a raid without major
consequence. In chapter 30, dealing with the conquest of Izmit, APZ claims
that this concluded the conquest of Kocaeli, Ak Yazı, Konurapa, and Bolu.
That would mean that the lands of the rich Ak Ova along the highway that
led east from Adapazar were now in Orhan’s control, while the expedition
narrated here represents a rounding out, by the final conquest of the territory
immediately to the south, along the road that leads east from Geyve and
the Sangarius.
KONURAPA poses a problem. The name does not appear in the official
list of place names in Turkey, the Meskun yerler kilavuzu, but a document of
the early sixteenth century lists a village of that name adjacent to the large
market town of Düzce.191 This is presumably the village now called Konur
Alp, situated near the ancient Prusias ad Hypium. This district, too, contains
many of the small wooden mosques attributed to Orhan. Bolu, on the other
hand, which lies some forty-five kilometers further east, contains impressive
monuments of Beyazit I, but nothing earlier.
Orhan Acquires Karesi: APZ 35–36
Aclan Bey, son of Karesi, died at this time, leaving a son called Dursun who
went over to Orhan. His brother Haci Ilbeg who had remained with his
father, was unpopular with the people. Dursun Bey proposed to divide his
lands with Orhan, whom he urged to take Balıkesir, Bergama, and Edremit,
leaving Kızılca Tuzla and Mahram for himself. Orhan then proceeded to
conquer much of Mysia. He took Ulubat, where he left the tekfur in place,
then moved from Gölbaşı to take Biluyuz and Ablayund. He arrived at
Kirmasti whose ruler was a woman called Kilemastorya. Orhan met her and
her brother Mihaliçi, brought gifts, and left her in place. The tekfur of Ulubat,
however, had not kept to the agreement he had made, and was finished off.
The next conquest was Balıkesir, from which the other son of Karasi fled,
taking refuge in Bergama. Orhan proclaimed that the people would have
security under his rule. The people of the region submitted, families
returned, and those with timars were left in place. This took place in 1335.
Orhan Completes Acquisition of Karesi,

Becomes Padishah: APZ 37
By taking the right of hutbe and sikke (being named in the Friday prayers
and striking coins) in Karesi, Orhan became padishah. Haci Ilbeg surren-
dered Bergama on terms and was sent to Bursa where he died two years
later. Orhan gave the whole of Karesi to his son Suleyman as timar.
The significance of Orhan’s new title of padishah is not evident. His coins
struck in Bursa show that he already had the right of sikke; he had claimed
hutbe since 1299 (APZ cap. 14).192 If the title padishah was really claimed at
his time, it presumably reflected his rule over more than one emirate.
The account of the Ottoman acquisition of Karesi presents irreconcilable
problems, which will be discussed next. The geography is real enough:
Balıkesir, Bergama, and Edremit were the main cities of the emirate, but
“Dursun bey” was unlikely to have been in a position to offer them all
together, for Bergama and Balıkesir were the headquarters, respectively, of
the maritime and inland branches of the emirate. The region of Kızılca
Tuzla and “Mehram” (presumably Behram, the ancient Assos) is implausibly
tiny, for the two places lie virtually side by side on the coast west of Edremit.
“Dursun” presumably uses these names to refer to the Troad, which was a
separate dependency under the ruler of Bergama.
APZ names only a few places among Orhan’s conquests in Karesi, and
these clustered around Lake Apollonia in easy reach of his capital Bursa
and on routes that led from there to the Dardanelles, Balıkesir, Bergama,
and Izmir. “Gölbaşı” is no doubt Başköy, which was for Fontanier in 1827 the
first stage west from Bursa, six leagues distant, at the edge of marshes that
stretched down to the lake.193 “Biluyuz” is presumably Balyoz in the hills
southeast of Lake Apolyont, while “Ablayund” is evidently Apollonia, a well-
known stronghold on a promontory jutting into the lake which bears its
name, twenty- five kilometers west of Bursa and well- fortified in the
Byzantine period194 (Fig. 1.36). The key point here is the powerful fortress
of Lopadion (Turkish ULUBAT) built by John Comnenus to defend the area
against Turkish attacks, and controlling a strategic river crossing.195 Osman
Fig. 1.36 Apollonia
threatened the city after his victory over the coalition of tekfurs at Dimboz
in “1302”; Orhan took it after the aborted restoration of its tekfur in the
present campaign. Yet how it managed to be in the hands of a tekfur at this
late stage is not at all obvious, for a Byzantine chronicle mentions its fall in
1327 (though not to whom).196
The main road west from Bursa led along the north shore of the lake where
it featured a substantial stone caravansary built in the fifteenth century. In
Ottoman times, the commercial center of the district was MIHALIÇ, where
the roads branched to the west and southwest. Locals participated in the
long-distance trade to such an extent that Lebas could report in 1844 that they
knew more about Bursa, Smyrna, and Constantinople than the interior of their
own district.197 The road that led to Balıkesir and Izmir gained considerable
importance in the seventeenth century and later with the rise of Izmir as a
major commercial center; a constant stream of travelers between there and
Istanbul passed through and left descriptions usually not very detailed since
this part of the road at least contained no large cities or impressive antique
remains.198 KIRMASTI (now called Mustafakemalpaşa) stands on the river
Rhyndacus, about ten kilometers southwest of the lake, on an alternative,
somewhat shorter though less practicable road southwest from Bursa. In the
early twentieth century it preserved the remains of a Byzantine castle and a
tomb supposedly of Lala Shahin, tutor of Orhan’s son Murat I.199
The name of Kirmasti’s female ruler “Kilemastorya” has plainly been made
up on the basis of the place name (just as Yalakonda from Yalova above),
while her brother “Mihaliçi” covers the name of the town Mihaliç (now
Karacabey) some five kilometers west of Ulubat. From here, the narrative
skips to the two capitals, Balıkesir and Bergama, omitting any mention of
the long and strategic coast—of the Marmara, the Hellespont, and the
northeastern Aegean—that Karesi controlled.
The following chapters, beginning with 39, describe the crossing to
Europe and the first Turkish conquests there.
How Reliable a Tradition?
The narrative of APZ is highly circumstantial, presenting a plausible-

looking sequence of events against the detailed background of a region with
an abundance of place names. Plausible or not, it needs to be tested to see
whether it is history, legend, or some combination, and whether it is inter-
nally consistent.200 Such an investigation may begin with the toponyms, to
determine whether the narrative is set in a real or imaginary background. A
story that unfolds in places that never existed, or were actually located far
from where they are set, would not inspire confidence.
In fact, this is a real landscape, and the toponyms are in their correct
places. Of the forty-five towns, forts, and villages that appear, fully forty-one
can be located positively or most plausibly in the region where they are set.
None of them is out of place. Only four cannot be identified, either because
they have disappeared or their name has been changed. Nothing here raises
suspicion. What is equally striking is the familiarity of the author with some
details of the land. Most notable is the allusion to the country around Çakır
Pınar, where the land really does open out, leaving the narrow crags behind.
Another local touch is mention of the little tower at Justinian’s bridge, a real
structure that would not be imagined if it were not seen. But presenting a
real landscape, though encouraging, does not validate the narrative, for
novels are frequently set in real landscapes, though their protagonists are
fictional.
Another step will help: is the setting appropriate to the time of Osman
and Orhan, or are their notable inconsistencies or anachronisms? Here, too,
the answer is positive. Of those forty-one identified toponyms, twenty-six have
evidence from the thirteenth/fourteenth centuries. This usually consists of
physical remains—the masonry of fortresses or finds of pottery—less often
of texts. That does not mean that the other fifteen sites are out of place
chronologically, only that they have not been investigated closely enough.
There are no places that definitely did not exist at this time. So, it can be
taken that the narrative is set in a real time and place.
Next, consistency: do the events make sense in their context and is their
sequence plausible? Does the narrative contradict itself or fail to agree with
whatever evidence is available from other, generally accepted, sources? Here
the answer is mixed.
Leaving aside for the moment the earliest events already discussed, but
beginning with the conquest of Yar Hisar, Bilecik, and İnegöl, a plausible
sequence of acquisition and expansion emerges. These strong points bring
Osman from the mountainous borderlands of Phrygia down to the hills and
plains of Bithynia, to a region much richer in agricultural or pastoral land
and better communications with the metropolitan centers of Bursa and
Iznik, the first real cities the Ottomans would encounter. From here, it was
logical to move on to Yenişehir with its broad plain and easier access to the
cities (Yenişehir was a good site for a headquarters if Osman had designs
against Byzantium). Logical, too, that the Byzantines should take alarm at a
movement plainly directed against them, and should organize the disparate
forces of the region in an effort to bring Osman’s ambitions to an end. His
victory at Koyunhisar was a decisive step, the first time he met a substantial
army, as opposed to picking off forts one by one, as often by stratagem as
not. He was now a formidable opponent, directly threatening the empire in
its Anatolian heartland. This sequence of events makes more sense than
APZ’s occasional chronological markers. If Osman could take Yarhisar in
“1288,” why did it take him another decade to overcome the eight kilome-
ters that separate it from Yenişehir?
Koyunhisar left the way open to Bursa, where the Turks, lacking siege
equipment, resorted to what was to prove a successful strategy of building
blocking forts with the ultimate aim of starving the cities into surrender. In
the case of Bursa, this would be a long—or long interrupted—process. They
now, in “1302” turned in the opposite direction, to overrun the crucial
Sangarius valley, heavily defended by Byzantium against enemies coming
from the east—not from the southwest. Success here was decisive, for
Osman could now send his forces over a pass and down toward Nicaea or
further north to Justinian’s bridge and the highway that led west to Izmit
and Constantinople.
The next event reported, in “1326,” seemingly after a gap of twenty years
or so, is the fall of Bursa, which succumbed after a long blockade, preceded
by the rather remote Adranos. With this success behind them, the Turks,
led now by Orhan, made a spectacular demonstration of force by penetrat-
ing to the suburbs of Constantinople and taking the steep fortress of Aydos.
Meanwhile, another contingent struck east into the remote but fertile region
of eastern Bithynia.
Like other goals of Osman, the fate of eastern Bithynia poses problems.
APZ (10) reports that soon after the conquest of Karacahisar, Osman pro-
posed an expedition against Tarakçı Yenicesi. Mihal advised him of the
route and told him that the rich land of Mudurnu would be easy to attack
because Samsa Çavuş and his followers were settled near there (as they had
been since the days of Ertuğrul). Samsa joined them after they crossed the
Sangarius and together they raided Göynük and Tarakçı Yenicesi and
reached Gölkalanoz (Gölpazar), whence they returned to Karacaşehir. They
took much loot but no captives because Osman wanted to treat the popula-
tion well, so that they might his subjects. Subsequently, after the successful
Sangarius campaign of 1304, Konur Alp took Düzpazar (Düzce) and
Akyazı, defeated the enemy at Uzunca Bel, and advanced as far as Bolu
(APZ 22). This was apparently a raid.
After the conquest of Bursa, APZ (25) reports that Konur Alp attacked
Akyazı, Mudurnu and Bolu as well as the land of Konurapa. The Anonymous
(5), though, attributes the conquest of Tarakçı Yenicesi, Göynük, and
Mudurnu to Orhan’s son Suleyman, as does APZ (34) who, however, places
this after the conquest of Iznik (or Izmit). He adds that these places had
surrendered willingly. Meanwhile, on the death of Konur Alp, Orhan had
assigned his lands to Suleyman (APZ 30).
Al-Uryan, who left Anatolia in 1333, reports the existence of an emirate
called Koynuk Hisar.201 When Ibn Battuta passed through, however, the
city, inhabited only by Greeks, was in Ottoman hands.202 It seems probable
that the three towns named together by the Anonymous (Yenice, Göynük,
and Mudurnu) and which form a coherent geographic unit, were all part of
an independent state adjacent to Ottoman territory and only taken by
Orhan. Osman’s lands did not include them, though they were suitable for
raiding. When APZ mentions the populations of these areas, they are made
up of infidels who submit without a fight and often become Muslim. Note,
though, that no tekfurs appear, perhaps suggesting that the writ of Byzantium
no longer ran in this distant province.
Bolu poses a special problem. Ibn Battuta, who passed through Mudurnu
and Bolu, does not indicate who controlled them, only that Kainuk was
in the territory of Orhan and that Gerede, the next major station east of
Bolu, had its own ruler. In fact, conquest of Bolu would have been a major
accomplishment because of the topography. The lowlands of the Sangarius
end about 25 km east of the river after which there is a climb to the fertile
plain of Düzce, the ancient Prusias ad Hypium, a place that was very pros-
perous in antiquity from export of timber.203 East of there, the road rises
steeply, a difficult track through dense forest, to reach the isolated basin of
Bolu, ringed by high mountains, at an altitude of 725m, compared with
Düzce’s 120m. Many travelers experienced the unpleasant passage, notably
James Baillie Fraser who in 1834 described the long and difficult forty mile
stage to Bolu through a dense forest of oak and beech—a thorny jungle—
with a river that constantly had to be crossed. 204
The next conquest poses a real problem. According to APZ, the united
forces of the Turks marched on Izmit, the natural goal of a force that
advanced down the Sangarius and past Lake Sapanca. But APZ narrates the
surrender of Izmit and Yalova, before proceeding on to the fate of Iznik,
which succumbed to an effective series of blockading forts. There seems to
be a serious confusion here. Izmit is not naturally connected with Yalova,
which is rather a port for Iznik, and the description of the fortifications, in
the valley and on the ridge, certainly suits the region of Iznik.205 In any case,
the three great Bithynian cities were now under Ottoman control and a new
era for Bithynia had begun.
The momentous conquest of the coast of Karesi, which made the crossing
into Europe practicable, poses different problems, for APZ gives its rulers
names that are unattested elsewhere and his date for the conquest—1335—
is manifestly wrong, for texts and coins show that Karesi was still function-
ing ten years later. The tradition evidently knew very little about Orhan’s
takeover of Karesi.
Despite some disquiet raised by the conquest of Iznik, Izmit, and Karesi,
the narrative from the capture of Bilecik to the domination of Bithynia
makes sense, showing an orderly progress where each stage lay the founda-
tion for the next. The same cannot be said of the account of the origins and
rise of the primordial Ottoman enterprise.
The notion that Ottoman history began in Söğüt, the residence of
Ertuğrul, is so well entrenched that there is no reason to doubt it. Even
though no remains from the time of Ertuğrul or Osman have been reported,
the role of the town is attested as early as the fifteenth century. The summer
and winter pastures assigned to Ertuğrul, however, inspire no such confi-
dence. For the summer, the sultan granted pastures at Ermeni Beli and
Domaniç. As already noted, these make little sense as summer pastures for
Söğüt because they are in the same climate zone and stand at very similar
altitudes. This does not mean that Ertuğrul had no sheep to pasture, just
that these were not the right places for them.
Osman supposedly had mixed relations with the tekfurs of İnegöl (hostile)
and Bilecik (friendly). It is perfectly plausible that Byzantium had outposts
in such places, for Bithynia was heavily fortified, especially its mountain
passes and major roads. The local commanders, often hostile to the imperial
government, were certainly capable of collaborating with the Turks.
By its universal use of the term tekfur, the Tradition shows a manifest
ignorance of its prime adversary. It presents the commander of each city as
autonomous, though capable of joining with rules of other cities in a joint
effort (as in the battle of Dimboz) and presents the tekfur of Istanbul as a
chief like the others. In fact, the Byzantine administrative system as known
in the Lascarid period was complex, with provincial and municipal gover-
nors subordinate to the emperor.206 The sense of a hierarchy—indeed, the
existence of a state headed by a supreme emperor who ruled a large terri-
tory—is completely lacking. Osman would probably have been aware of the
nature of his adversaries, as Orhan certainly was, but the Tradition has dis-
pensed with the details.
Karacahisar poses a special problem because of its location on the road
from Kütahya, headquarters of ever-hostile Germiyan, to Eskişehir; this
does not seem capable of resolution. Finally, the collaboration with the
Christian commander Köse Mihal, seems to involve a figure whose very
existence is more than doubtful. In sum, the traditions about Ottoman ori-
gins are dubious or tendentious. APZ and his sources seem to have known
very little about them and compensated by incorporating stories that had
the ring of plausibility. For this period, there is only the penumbra of shad-
owy people and events moving around a real landscape. The situation only
changes when Osman moves out of the confines of his first home to engage
with a broader world. From the conquest of Bilecik, Yarhisar and İnegöl, the
scene becomes more plausible and some confidence may be placed in the
tradition, at least until the capture of Iznik and Izmit.
The physical environment provides a context for the traditions about
Osman and Orhan and to some extent allows them to be verified or rejected.
The physical context is in itself informative: a rough, hilly, or mountainous
country with few natural resources and no large plains makes it unsuitable
for a nomadic society or for a powerful centrally organized state. It would
also have been a land its neighbors wouldn’t covet. A country that could not
support a large population, it never had any cities even in the most
flourishing ages of Rome. Osman’s realm was evidently quite poor until it
spread to the fertile coastal plains. Its location on an important cross-country
route should have made a difference, but the sources are silent about trade
though not about the raids that swept along the roads.
This country also (and this seems not to have been noticed) provides
a real context for the endowment of Jibrail b. Jaja, whose lands were the
direct neighbors of Ertuğrul and the young Osman. His territories around
Eskişehir—the most suitable for flocks and herds of nomads—were given
over to agriculture, suggesting that decades of stability under the Seljuks
had tamed a land afflicted by the Turcomans of the frontier. More
importantly, the location of ibn Jaja’s lands brings the Mongols into the
picture at an early date, attested by real documentation not an oral tradition.
Their presence in the Homeland raises questions about the relation of the
incipient Ottoman enterprise to the all-powerful Ilkhanids and for that
matter to the newly established Germiyan. These questions will be
approached in Chapter 6, but any answer to them will involve a good
dose of speculation.
Notes
1. Kafadar 1995, 96–104 puts the sources into their historical context; for the com-
plexities of the sources see the discussions of Inalcık 1962 and Menage 1962; and
for Yahşı Fakih, Menage 1963, cf. Haşim Şahin “Yahşı Fakih” in IA (2013); and
Menage 1964 for Neşri.
2. Edited and translated by Kemal Silay; see Bibliography for details.
3. Edited and translated by Irene Melikoff-Sayar as Le destan d’Umur Pacha;
henceforth referred to as Destan.
4. For editions of APZ, see the Bibliography at the end of this book.
5. See Lindner 2007, 15–34.
6. “APZ” will be used for “Aşıkpaşazade”; “Homeland” denotes the lands occupied by
Osman before the battle of Bapheus in 1302.
7. See the article “Ertoghrul” by V. Menage (1965) and “Ertuğrul Gazi” by Fahmattin
Başar in IA (1995) and note that the coin inscribed “Osman ibn Ertuğrul” is
of dubious authenticity: see p. 142f. On the other hand, he is mentioned as
grandfather of Orhan in a foundation document of 1361: see Beldiceanu 1967,
131–4 with n. 1.
8. Neşri I.74f.; the place still existed in 1858 when Mordtmann (1925, 549) passed
through it on his way from Eskişehir to Söğüt.
9. The name no longer appears on modern maps, the traditional “Dog’s Nose”
apparently not thought sufficiently respectable for modern taste. It was changed
in the 1930s to a more innocuous Ilk Burun, “First Point” and so appears on the
map in MAMA V (also, p. xviii). Later it became an even blander Ulu Dere,
“Great Valley,” the name it still bears. Such changes, which sacrifice history to
current tastes or bureaucratic convenience, were frequently denounced by
Louis Robert, e.g., 1977, 57–63.
10. For the location, resources, and communications of Söğüt, Lindner 2007, 35–56
is indispensable.
11. Evliya Çelebi IV.206f.; cf. Haji Kalfa 702. Later descriptions: Leake 1824, 15f.,
Kinneir 1818, 33–5, and Keppel 1831, 174ff.
12. Cuinet 1894 IV.179.
13. This geographic classification is from the admirable work of von der Goltz
1896, 294; cf. in much more detail Geyer, “Donnees geographiques” in Geyer-
Lefort 2003, 23–40. For the environment and its changes in historic times, see
the articles in Geyer-Lefort 2003, 153–205 and 535–45.
14. For the roads see Haji Kalfa 702, Taeschner 1924, I.77–151 especially 123f., with
references to sources of the fifteenth century and later and Lindner 2007, 45–50,
54–6. For a comprehensive view, see Geyer-Lefort 2003, 461–72.
15. The first quote is from Leake 1824, 14, the second from Kinneir 1818, 33; cf.
Fellows 1829, 121.
16. See the itinerary of General Koehler in Leake 1824, 143; cf. Keppel 1831, 174–8
and Fellows 1839, 123f.
17. Evliya IV.206f. The tomb is discussed by Ayverdi 1966, 198–200.
18. For the campaign of Tamerlane, see Alexandrescu-Dersca 1977, 80–5.
19. Ahmedi 48; Chalcocondyles 11; I am indebted here to the observations of
Mr. David Barchard.
20. Söğüt has been identified with the Byzantine village Sagoudaous, mentioned by
Anna Comnena XIV.1. It depends, however, only on the resemblance of name,
and is to be rejected; see the discussion on p. 135.
21. Günyarık existed as early as the fifteenth century, when it appears in a list of
pious foundations in the province: Refik 1924, 133. Dating of the second period
of the church is suggested by surviving fragments of sculpture, perhaps of a
ciborium arch. For examples in a similar style see Grabar 1978, plates 85b, 105,
113b, 123b, 139c, 139d, of the thirteenth and fourteenth centuries, the latter
somewhat flatter than the present examples. My thanks to Prof. Cyril Mango
for his comments.
22. For the route of the crusaders, see Runciman 1953, 180–6; cf. the usually
neglected comments of von der Goltz 1896, 456f., who saw the landscape
through the practiced eyes of a general, and suggested a different site for the
battle of Dorylaeum from that usually accepted.
23. Here, as elsewhere, I take elevations from the excellent map of von Diest 1898.
24. For Armenokastron, mentioned only by Anna Comnena XIV.iii, 6, see Wittek
1935, 36.
25. For this route, see Taeschner 1924, 199f., and the examples of its use by Newberie,
who traveled in 1581/2, in Purchas 1625, 1419, and Tournefort 1717, 335–9.
26. Taeschner 1924, 97–100, 122–34; for what follows, see Dernschwam 1923,
163–6 = 2003, 126.
27. See, for example, MacFarlane 1850, 317–22, Humann and Puchstein 1890, 12f.,
and von der Goltz 1896, 142–8 with illustrations.
28. Odo of Deuil 104, discussed by Tomaschek 1891, 89f.
29. Browne in Walpole 1820, 113.
30. Bertrandon 1892, 129f.
31. Cinnamus 38, a text that presents problems beyond the scope of the present
discussion.
32. Belon 1555, 359–61; Evliya Çelebi XIII.44f.
33. Lucas 1712, 114f., Olivier 1807, 502f.
34. Kinneir 1818, 239–41; Texier 1892, 392; Mordtmann 1925, 61–7.
35. See Texier 1892, 302f., Keppel 1831, 389f., and MacFarlane 1850, 238–52, with
his usual vivid description of local conditions.
36. See n. 22. I cannot understand the figures for heights and vertical range given by
Lindner 1983, 20.
37. tekfur, a term derived from the Armenian, is used to denote the infidel com-
mander of a city of whatever rank or function; see p. 87.
38. Discussed in Chapter 3, p. 136.
39. Evliya XIII.43f. For the Ottoman buildings, see Ayverdi 1966, 500 and 1974,
292–304.
40. See Kinneir 1818, 243; Keppel 1831, 391 (this quote); MacFarlane 1850, 232–8;
Mordtmann 1925, 68f.; Humann and Puchstein 1890, 11f.
41. Mordtmann 1925, 69.
42. See the colorful description of MacFarlane 1850, 322–45.
43. For masonry of this period, see Foss 1982 and Foss and Winfield 1986, 150–9.
44. Bilecik is not mentioned in Byzantine sources, though modern writers have
identified it with the Belokome of Pachymeres XI.21 (4.453). This is to be
rejected: see the discussion in Chapter 3, p. 135f.
45. See the detailed discussion of Ayverdi 1966, 29–40; cf. Kuran 1968, 68f. Note
that the adjacent “tomb of Mal Hatun” is a more recent structure.
46. Ayverdi 1966, 36–40.
47. Dates in quotation marks are those given by APZ that cannot be verified. They
are used here as marking the sequence of events without implying that they are
correct.
48. Humann and Puchstein 1890, 11f.; Ayverdi 1966, 5.
49. Kaplanoğlu 2000, 23 mentions fortifications at a village called Süpürtü 3–4 kilo-
meters from Kulaca, whose inhabitants reported that the fortress was formerly
called Kulaca.
50. Kalanos also appears as Falanos, mentioned in a document of the sixteenth

century as being in the district of Göynük: Beldiceanu 2003, 365 n. 108. For
Alınca, see ibid. 365 n. 113.
51. For its location, between İnegöl and Pazaryeri, see Beldiceanu 2003, 365 n. 112.
52. The routes of this district will be considered next in connection with İnönü.
53. Remains described and illustrated in TIB 7, 287.
54. See earlier, n. 42.
55. It might, for example, be one of the forts which Manuel Comnenus provisioned
in 1175 before the reconstruction of Dorylaeum: see Cinnamus 294.
56. Neşri I.64–9.
57. Neşri I.74–7.
58. Neşri 71–3; Haci Bektaş: Taeschner 1928, 101 note 1.
59. APZ cap. 8. The recognition of Osman by the Seljuks was the subject of docu-
ments now recognized as modern forgeries: see Beldiceanu 1967, 59–77.
60. APZ caps. 6, 8, 9.
61. APZ cap. 9.
62. APZ cap. 14, but note that the Anonymous p. 12 gives the date as 689 (1290).
See the discussion by Danişmend 1947, 5f.
63. APZ cap. 21; for the Çavdar Tatars, about whom very little is certain, see the
exemplary study by Naumann 1985, 284–8.
64. APZ cap. 21; the date is uncertain: APZ places it immediately after a campaign
of 1305 along the Sangarius.
65. Document: Refik 1924, 130, 132. Dernschwam 1923, 169; Haci Kalfa 702f.
66. Von der Goltz 1896, 179f.; location of Dorylaeum: Radet 1895, 491–513, largely
refuted by Körte 1897, 388–94; inscriptions: MAMA V.115–17.
67. For an introduction see Parman 2001, with a plan and new photographs. Progress
reports are published most years in the Kazı Sonuçları Toplantısı: the most
comprehensive so far is in the 24th (2001) volume of that series: 24.2, 69–80.
68. Körte 1895, 14f.; 1897, 388f.; MAMA V plates 4 and 5.
69. Cinnamus 294.
70. Ayverdi 1966, 13. The titles attributed to Osman raise the suspicion that the
document (apparently unpublished) is a fabrication based on the mentions of
Karacahisar in the sources.
71. APZ cap. 15.
72. RCEA xiii, 5080; for the history of Ankara in this period, see Wittek 1932, 340–53.
73. See the descriptions of von Diest 1898, 27f. and Haspels 1971, index s.v.
Kumbet, Oines.
74. The identification may need to be reconsidered; Beldiceanu 2003, 362 locates
Oynaş at a place called Oynuş in the region of Kütahya, while Kaplanoğlu 2000,
25 identifies it with Saruhan immediately north of Tavşanlı—a better location
for the narrative—but gives no reason for the identification.
75. Haci Kalfa 702; Otter 1748, I, 50f.

76. Leake 1824, 142f.
77. Keppel 1831, 179f.; cf. Fellows 1839 who made the ascent and found the caves
“an excellent substitute for a castle.”
78. See TIB 7, 281.
79. Huart 1897, 38 (incomplete transcription only).
80. Kuran 1968, 70; cf. Kızıltan 1958, 106–10 and note that the mosque is absent
from the masterful surveys of Ayverdi, presumably because he did not consider
it an Ottoman work.
81. Varlik 1974, 57.
82. APZ cap. 30.
83. For Dorylaeum in this period, see Foss 1996b.
84. Al-Harawi cap. 58 (ed. Meri 152f.).
85. Ibn Said ap. Cahen 1974, 44. Cahen gives the name as Sultanbuli, which I pre-
sume appears in the Arabic text—not reproduced—as an error for Sultanyuki,
from which it differs only by one dot and a slanting line, viz.:
‫ سلطان بويل‬and ‫سلطان يويك‬. Al-Harawi has ‫ سلطان ويك‬which, too, is probably to be
emended slightly to be equivalent to the Sultanyuki of al-Umari 39.
86. RCEA xii, 4596,
87. The will appears in two versions, Mongolian and Arabic; both are edited, in a
most exemplary fashion, with Turkish translation, introduction and notes, by
A. Temir 1959.
88. Temir 1959, 151–219.
89. Temir 1959, 61–4, 127–9.
90. J. H. Mordtmann, Fr. Taeschner, “Eskişehir”.
91. APZ 29; this anomalous use of the name Sultanönü may perhaps be taken to
denote a district, while “Eskişehir” meant the city.
92. Al-Umari 32/350.
93. Another passage of the same author confirms the location indicating that the
land of Sultanyuki was situated adjacent to the principality of Kastamonu and
the territory of the house of Jenghiz Khan, that is, the land ruled directly by the
Mongols: Al-Umari 39.
94. Ibn Battuta II, 324, 342.
95. See Beldiceanu-Steinherr 1967, 99 n. 32.
96. Otto-Dorn 1941, 10, 69; the building does not survive.
97. Refik 1924, 135, 141.
98. Evliya Çelebi IV, 207.
99. Haji Kalfa 701; the tomb apparently still exists: Eskişehir Il Yıllığı 96.
100. See, e.g., Mordtmann 1925, 549; the best description seems to be that of von
der Goltz 1896, 173.
101. Temir 1959, 61–4, 127–9.
102. Similar lists that suggest the use of a common chancery style to describe such
properties may be found in the slightly earlier vakf documents published by
Turan 1958, 112, 141. Yet, even if these phrases are all formulaic, the specific
items that follow certainly constitute concrete evidence.
103. For silk production and trade in Seljuk and Ottoman Anatolia, see H. Inalcık,
“Harir”.
104. See the map at the end of Erdmann 1961, with the references there to his
meticulous descriptions.
105. Some of these village names appear in the vakf document of the late fifteenth
century in slightly different forms, Eğrigöz and Söğüt Öyüğü: Refik 1924, 133,
135, 140; the differences presumably arise from the use of Turkish in this and
Arabic in the will of Ibn Jaja. For Beş Kardeş, see Leake 1824, 17 and, in more
detail, with photographs, MAMA V, xviiif.
106. Refik 1924.
107. For It Burnu, see p. 10.
108. See the photographs in MAMA V, plates 2 and 3, and the descriptions of
Dernschwam 1923, 170, who seems in 1555 to have missed Eskişehir alto-
gether, and of Kinneir 1818, 35f.
109. Ahmedi 157; the date is discussed by Wittek 1932, 351ff. and by Beldiceanu
1965, 444f., who would move it back to the last years of Orhan; cf. also Inalcık
1965, 154ff.
110. Refik 1924 passim; Osman and Orhan: ibid., 134.
111. See n. 123.
112. APZ cap. 10.
113. See the long article “Mihaloğulları” of M. Tayyib Gökbilgin in IA (2005) and
the first pages of Gazimihal 1958; the comments of Ayverdi 1966, 150f., 1972,
170f. are, as usual, clear and sensible. Discussion of the subject seems inevita-
bly to suppose that any prominent figure named Mihal had some connection
with Köse Mihal, without reflecting that the name Michael was extremely
common among Byzantines and that any number of converts who, for what-
ever reason, did not adopt a Muslim name may have borne it.
114. Mihal is described as Sahib ul-khayr, dafi` ul-dayr and a`dal ul-umera; see
Ayverdi 1972, 170f.
115. Gazimihal 1958, 129f.
116. Refik 1924, 137. The text actually reads “Mihallardagi Gömelede,” corrected by
the editor to give better sense.
117. The same objection could be raised regarding Lowry’s (2003, 56–67) discus-
sion of a document of 1390 (known only in a later copy) which grants land,
privileges and extravagant titles to a certain Ali beg son of Mihal beg. Ali was
being rewarded for his services in the battle of Kosovo in 1389. Lowry pre-
sumes that the Mihal of this document was Köse Mihal, companion of Osman,
who is last mentioned in connection with the capture of Bursa in 1326. This
would make for an implausibly long generation. This Mihal is named without
any indication that he was one of the great heroes of his age.
118. Note that Imber 1993, 67f. and 1994, 131f. also rejects the historicity of Mihal,
by a different chain of reasoning. Kafadar 1995, 26, 127 accepts Mihal as an
historical figure. Kiprovska 2013 offers the most comprehensive defense of
Mihal as a real figure, a Byzantine commander alienated from the government
who joined the Turks, a general circumstance described by Pachymeres. She
cites all the sources and previous discussions, placing great importance on the
association of Mihal with Harmankaya, and concludes (263): “the explicit evi-
dence for the family’s hereditary command of the infantry troops of the area
strongly implies that this situation originates in the nascent years of the
Ottoman state with the forefather of the family—Köse Mihal.” For the military
organization of the region of Eskişehir under the early Ottomans, with evi-
dence beginning in the reign of Orhan and discussion of a document of 1466
that associates Harmankaya with Mihal bey, see Doğru 2005, 107–16.
119. Kiprovska 2013, 266, illust. 2.
120. See Kiprovska 2013, 261 n. 72.
121. Gazimihal 1958, 128f.
122. Refik 1924, 134.
123. According to a document quoted by Beldiceanu 2003, 360n58, the zaviye was
in Eskişehir, which of course is not on or near the Sangarius. Either there were
two places of the same name or the tradition is seriously confused.
124. The same name was applied, for example, to the Roman obelisk outside Nicaea:
Pococke 1745, 123; and the bridge of Justinian over the Sangarius near
Adapazar is still called Beşköprü, “Five Bridge” from its five arches.
125. Von Diest and Anton 1895, 16.
126. Sahin 1981, 32, von Diest and Anton 1895, 15.
127. Ibn Battuta 325.
128. See the description of the French engineer Pouillaude quoted by von Diest and
Anton 1895, 9f., cf. 13.
129. Von Diest and Anton 1895, 15f., with 13f. for the general characteristics of
the valley.
131. Evliya Çelebi IV.172.
132. See Kinneir 1818, 264f. for the town and its surroundings.
133. Von der Goltz 1896, 260–9.
134. Kinneir 1818, 264–71 (Taraklı-Bolu), von der Goltz 1896, 255–8.
135. Ibn Battuta 329, there called Kaynuk.
136. Evliya Çelebi IV.172.
137. See the description and comments of Ayverdi 1966, 145–8.
138. Ayverdi 1966, 10–12.
1 39. Ibn Battuta 330–4.

140. Von Diest and Anton 1895, 111f.
141. For the road, see French 1981.
142. Ayverdi 1966, 170f.
143. Ayverdi 1966, 201–4.
144. Giros 2003, 217.
145. Lefort 2003, 104; Grelois 2003, 124f.
146. Grelois 2003, 123f.
147. Evliya 13, 42f.
148. Ayverdi 1966, 15f., 205–16; the dervish lodge is the building that drew

Dernschwam’s attention. There seems to be little evidence for attributing the
bath to Osman.
149. Pococke 1743, III.121.
150. MacFarlane 1850, I.220f., Texier 1862, 144; Mordtmann 1925, 69;

Sarman 2001, 5.
151. Geyer-Lefort 2003, 25f.
152. Fortifications first reported by Kaplanoğlu 2000, 26.
153. Hamilton 1842, 84, 90; Texier 1862, 142f.; Giros 2003, 224. For Kirmasti,
see p. 82f.
154. Beldiceanu 2003, 369 n. 137.
155. Evliya 13, 43; MacFarlane 1850, I.103; Giros 2003, 221.
156. Hasluck 1906, 300; Giros 2003, 222f. (with plan).
157. But see the objections of Beldiceanu-Steinherr 2003, 371.
158. Hasluck 1910, 78–83; cf. Foss and Winfield 1986, 147f.
159. For the second, now called Balabancık Hisarı, see Ayverdi 1966, 8f. Its undis-
tinguished masonry of mortared rubble and occasional brick courses could
suit this period.
160. Beldiceanu-Steinherr 2003, 369 n. 140.
161. Evliya 4, 205.
162. von der Goltz 1896, 122, see his map p. 114; not on the Turkish 1:200,000 map.
163. Foss and Winfield 1986, 156; Giros 2003, 217f., with sketch plan.
164. von der Goltz 1896, 121f., 403; Giros 219; Yıldırım 2006, 42–5 suggests a date
of the twelfth–fourteenth century on the basis of the rough fieldstone facing
with little brick.
165. Discovery and discussion: Foss 1990, 166–73; cf. Giros 2003, 217–19, with
plan; detailed description: Yıldırım 2006, 46–67.
166. von der Goltz 1896, 10; Evliya 4, 174. Description and dating: Yıldırım 2006,
72–6. The correct form of this name, which appears under various spellings,
seems to be Kara Çepiş, “black yearling goat,” not surprising in a country that
has Goat, Sheep and Shepherd castles, and perhaps reflecting some local
folk tale.
167. Pachymeres I.149.
1 68. Evliya 4, 173.

169. Beldiceanu-Steinherr 2003, 370 with n. 141.
170. It is hard to know where to put the state of “Qawiya,” ruled by Murad al-din
Hamza and sometimes associated with Geyve. See the discussion in Chapter 7.
171. Foss 1990, 176. An alternative identification might be Büyük Kale, on the river
about two kilometres to the south: Giros 2003, 220.
172. See the comprehensive discussion by Whitby 1985.
173. The river was heavily fortified: for surviving structures between the gorge and
Adapazar, see Foss 1990, 176, and Mağara and Sefiler, both discussed in more
detail by Yıldırım 2006, 84–7, 95f., and between Adapazar and the sea:
Yıldırım 2006, 88–94 (Harmantepe, also called Söğütlü: Giros 2003, 220).
These forts can all be associated with the late thirteenth century strengthening
of the Byzantine frontier.
For Ak Yaziı, see von Diest 1898.
174. Texier 1862, 91; von der Goltz 1896, 401–10, esp. 408; Giros 2003, 216.
175. Osman seems to be a bit confused here: Bay Hoca fell in the battle with the tekfur
of Inegol at the Ermeni Pass at the very beginning of Osman’s wars of expansion
(APZ 3); the nephew who was killed at Dimboz was Aydoğdu son of Gündüz
(APZ 17).
176. See the Appendix in Chapter 6.
177. Giros 2003, 212, and, for its history in the Byzantine period, Bondoux 2003, 391f.
178. Evliya 8, 142f.
179. See the Appendix in Chapter 6.
180. Von der Goltz 1896, 320–69.
181. Note the remarks of MacFarlane 1850.II.448 and Dörner 1941, 9–11.
182. Kandıra: Ayverdi 1966, 130f.; Akça Koca: ibid. 131f.; wooden mosques:

ibid. 120–33.
183. See Aygil-Özcan 2012, 69–82 for the foundation and early history of the
monastery.
184. Janin 1964, 451f.
185. Çelik 2011, 45–54 with excellent illustrations.
186. Evliya 3, 69.
187. Yalakabad: Evliya 13, 42; Çoban Kale surveyed and discussed: Foss 1996, 63–8;
cf. Giros 2003, 215.
188. Anonymous 14–16; this passage is translated from a slightly different version
by Inalcık (1993) 83f. Note that the extract from Neşri translated on 85f. relates
to Osman’s earlier attack on Nicaea in 1306–7.
189. Inalcık 1993, 87, 89.
190. Giros 2003, 215f.
191. See: http://www.kurucasile.gen.tr/1530-yilinda-bolu-sancagi-koy-adlari-2/
192. For Orhan’s coins, see p. 143f.
193. Fontanier 1829, II.95f.
194. Hasluck 1906, 67–73.

195. See above, p. 63.
196. See p. 122.
197. Lebas 1845, 39. See Hasluck 1906, 68–87 for the country, its remains and its
roads; cf. Munro and Anthony 1897 and for Mihaliç Texier p. 156.
198. Yerasimos 1991, 66f.; the best description is probably that of van Egmont
1759, 174–89.
199. Texier 1862, 143; Hasluck 1906, 74f.
200. These remarks deal only with internal consistency of APZ; confrontation of his
narrative with the Byzantine will follow in a later section, “Reconciling the
Accounts.”
201. Al-Umari 340.
203. For the geography of this region, especially of the plain of Prusias/Düzce, with
extensive citations of the travelers, see Robert 1980, 11–106.
204. Ibn Battuta 456–60.
205. See further discussion, p. 138.
206. See Angold 1975, 250–75.
2
The View from Byzantium
In order to put the rise of the Ottomans into a context—and ideally to check
the Turkish sources against those from outside their realm, in the hope of
finding confirmation or inconsistencies—it may be useful to consider the
entire decline of Byzantine Asia Minor in the time of Osman and Orhan.1
Byzantine sources are indispensable for their details and chronology. They
begin with George Pachymeres (1242–c.1310), born in Nicaea, educated in
Constantinople, and recognized as a leading intellectual of his day. A cleric—
apparently a deacon—he served in the high legal offices of the patriarchate.
His massive and enormously detailed history covers the period 1260 to
1308 and pays special attention to Asia Minor as he chronicles the decline of
imperial power there. Often critical of government policy and corruption,
he presents his material objectively and provides all we know (in a literary
tradition) of Osman particularly and the Turkish invasions in general from
Osman’s first appearance on the scene at the battle of Bapheus in 1302 until
his defeat by the Mongols in 1307. Pachymeres presents the half-century
from 1258 to 1307 in considerable detail in a virtually impenetrable preten-
tious classicizing style.2
Pachymeres’ work was taken up by Nicephorus Gregoras (c.1290–1360),
who carried the story down to 1358. A highly educated polymath and a
teacher much involved in the ecclesiastical controversies of the day, he was
entrusted with important commissions by the reigning emperors, all of
whom he knew. Gregoras presents the 1340s in particular detail and devotes
much space to theological controversies. He pays much less attention to Asia
Minor than Pachymeres, but provides a basic outline of events in a relatively
clear classicizing Greek.3
The latest period covered here is the subject of the memoirs of an
emperor, John Cantacuzene (1295–1383), who reigned from 1347 to 1354
and was responsible for inviting the Ottomans to cross into Europe. Well-
connected by birth, he moved in the highest circles until triumphing in a
civil war that made him emperor. Forced from power by another civil war,
he devoted much of his long retirement (as a monk) to producing a detailed
account of his career and accomplishments, much of it notably self-serving.
DOI: 10.1093/oso/9780198865438.003.0003
Reigning at a time of growing Turkish power, he had close relations with

Orhan and with Umur emir of Aydın, providing considerable information
about the empire’s relations with the Turkish rulers of Anatolia.4
If Aşıkpaşazade (APZ)’s work is like a song of triumph ever increasing in
tempo, Pachymeres produces an unremitting dirge, for his story is one of
loss, corruption, incompetence, and betrayal, as the empire tried in vain to
maintain its grip on its ever-diminishing Asian territories. To understand
the process, a brief glimpse backward at the confrontation between Byzantium
and the Turks will set the stage.
Everything followed from the battle of Manzikert in 1071, when the
Turkish leader Alp Arslan defeated the army of the overextended Byzantine
empire. Its emperor was captured and the succeeding civil wars left its fron-
tiers open to an unlimited advance that saw a Turkish sultanate established
in Nicaea and other states springing up on the Aegean seaboard. Thanks to
help from the First Crusade, the emperor Alexius (1081–1118) managed to
push the Turks back onto the Anatolian plateau. His successors John and
Manuel regained control of the coasts of Asia Minor and achieved an equi-
librium with the Seljuks of Konya. They could reach a settlement with an
organized state, but not with the scattered and seemingly innumerable
Turkish tribes, which defied all authority and raided deep into Byzantine
territory. The empire responded with a defensive system based on fortresses
that they built throughout their lands. These served as bases for the army,
refuges for the population, and a deterrent to attackers. Notable among
them was the network of the Neocastra established by Manuel in central-
western Anatolia between 1162 and 1173.5
No sooner was this done than Manuel, in an effort to smash the Seljuks,
met disaster at Myriokephalon in Phrygia in 1176. His ambushed army was
destroyed and frontier defenses collapsed, leaving the country open to
Seljuks and tribes. The effects were felt quickly: the last outposts of Byzantium
on the central plateau, the powerful fortresses of Dorylaeum and Cotyaeum,
fell respectively before 1180 and in 1182 to become Sultanyuki/Eskişehir
and Kütahya, valuable frontier posts for the Seljuks. The southwest coastal
region of Lycia was also lost by 1191, but the fate of the strategic outpost of
Laodicea was more complicated.6 This fortified city was of special impor-
tance because it controlled access from the plateau to the fertile Maeander
valley. Turks captured it in 1193, but lost it twelve years later to Theodore
Lascaris, who in 1206 ceded it to the renegade Manuel Mavrozomes, father-
in-law of the Seljuk sultan. He held it until 1230, when it reverted to the
sultan. In 1256, however, at a time of Seljuk weakness, Byzantium once
The View from Byzantium 101
again occupied Laodicea, only to lose it to Turcomans, who established the

first state based on tribes and formally recognized by their Mongol overlords
around 1261.
The next disaster was the capture of Constantinople by the Fourth
Crusade in 1204. The Byzantine realm broke up completely, with small
independent states springing up in Asia Minor, at Nicaea, Philadelphia,
Priene, Laodicea, Pontic Heraclea, and Trebizond. Of these, the Bithynian
domain of Lascaris became the empire of Nicaea by subduing all the rest
except Trebizond, an empire of its own. Not surprisingly, these years saw
impressive Seljuk gains, notably the major ports of Attalea on the
Mediterranean in 1207 and Sinope on the Black Sea in 1214. That was taken
from Trebizond, for by then Seljuks and Nicaeans were at peace after a sig-
nificant Nicaean victory at Antioch on the Maeander in 1211, which
brought a half-century of relative peace and stability to western Asia Minor.
The Lascarids took advantage of this time to erect or restore fortresses and
city walls throughout their domain.
The final decisive events—the Mongol defeat of the Seljuks and the recap-
ture of Constantinople—did not look like disasters for Byzantium. In 1243,
the Mongols, newly arrived on the scene, decisively crushed the Seljuks.
This seriously weakened the Seljuk state, who soon lost control of the frontier
tribes, now more able to encroach on Byzantine territories. Finally, in 1261,
the Nicene empire regained Constantinople, a victory that was to prove
ominous for Asia Minor, since the restored empire now had to deal with
numerous adversaries in the west. For that, they tended to transfer troops
from Anatolia, further weakening the defenses of that frontier.
In 1256, Michael Palaeologus, then governing Bithynia but fearing for his
future under a new emperor Theodore II, crossed the Sangarius and fled to
the Turks. Here, “he came to the dwellings of the Turcomans. This is a people
who occupy the furthest boundaries of the Persians [Seljuks] and feel
implacable hatred for the Romans [Byzantines], delight in plundering them,
and rejoice in booty from wars; this especially at the time when Persian
affairs were agitated and thrown into confusion by Tatar [Mongol] attacks.”7
Pachymeres set the scene for what follows by explaining that the eastern
frontier was protected by mountains and by the holders of land grants who
received generous subsidies and fought well as long as they were paid; but
when their pay was cut, whether from corruption or war expenses, they
were capable of moving away or even joining the enemy at a time when
many frontier fighters were being transferred to the wars in Europe. Those
who surrendered voluntarily became guides and allies, allowing the Turks
to make inroads into imperial territory and occupy strongholds there.8 This
was at a time when many Turks, fleeing from the Mongols, were moving
into defensible positions in the mountains. The situation was made worse
by the impossibility of dealing with the nomads. The Seljuk government,
such as it was, couldn’t hold them back and their insidious infiltration was
hard for Byzantium to control. Treaties made with Konya were of no help,
and agreements with the tribes, who would violate them at a moment’s
notice, proved useless.
Michael Palaeologus, as emperor (1258–1282), took a serious interest in
the Anatolian frontiers. One of his first actions, early in 1259, was a cam-
paign that brought the whole army in a show of force to Philadelphia, where
he inspected the frontier fortifications, installed garrisons and made lavish
gifts to the defenders.9 Yet keeping the frontier warriors happy was difficult,
for three years later, the peasants of the strategic mountain pass above
Nicaea (a place called Trikokkia is specifically mentioned) revolted.10 They
followed a blind boy who claimed to be the young emperor John IV Lascaris,
deposed and blinded by Michael. When the imperial forces moved against
them, the rebels, who knew the country and occupied the heights, held
them off in classic guerrilla warfare. They finally accepted an amnesty which
involved good treatment for those who surrendered, but harsh punishments
for the resisters. The punishments, however, could not be excessive as these
peasants were too essential for defending the frontier. The government also
had to be wary in dealing with the population of Anatolia, where support
for the patriarch Arsenius, whom Michael had deposed for his condemnation
of the emperor’s usurpation, was widespread. Adherents of the patriarch,
called Arsenites, maintained their opposition to the imperial church estab-
lishment until 1310. The revolt provided an ominous example for the future.
For forty years, from the 1260s to 1304, the main Byzantine effort was
directed at the Aegean region—the rich, fertile, and strategic plains of the
Cayster, Hermus, and Maeander lands that generated far greater wealth
than Bithynia, and were the places where the most successful Turkish states
were first established.11 The upper Maeander valley and adjacent regions
were the goals of expeditions that Michael led in 1260 and 1261 in response
to Turkish attacks.12
In 1264, the emperor’s brother, the despot John, led a campaign based on
the Maeander; he secured that as well as the Cayster and provided for the
soldiers of Magedon in northern Lydia. He transferred some of these skilled
archers to Europe with good pay and consolidated the position of those
who remained by coming to an arrangement with the Turks that would put
limits on the areas they could exploit for grazing.13 Nevertheless, he could
not recover the coastal regions of Strobilos and Tracheia south of the
Maeander. Likewise, the Byzantine position beyond the Sangarius in eastern
Bithynia and Paphlagonia was collapsing because of high taxes, engendered
partly by imperial extravagance and by a desire to keep the locals too poor
to revolt. Hostility to the government impelled many poor farmers, espe-
cially those on the frontier, to join the Turks who accepted those who joined
voluntarily. They collaborated in harassing the loyalist population, with the
result that many emigrated, leaving room for the Turks to infiltrate. The
emperor did nothing in the belief that the districts were near at hand and
could easily be recaptured; he was devoting his attention to the west.14
When John returned in 1267, he found a desperate situation. The Turks
were overrunning imperial territory in the absence of adequate defense:
populations had fled from the once-prosperous Maeander valley, while the
Cayster region, the mountain pass of the Neocastra, Abala, and Magedon
were all under attack. Coastal Caria, Byzantine until recently, had become
the base for enemy pirates. Paphlagonia beyond the Sangarius was virtually
depopulated, with only fortified towns on the coast surviving (they could
no longer be reached overland).15
A curious incident in Nicaea gives some insight into the jittery mentality
of the time. On February 23, 1265, in the middle of the morning, a rumor
suddenly spread that a great force of Mongols had attacked Nicaea, slaugh-
tering the guards at the city gate, and killing everyone they met as they
entered. The population panicked: a crowd gathered and started rushing
around, while others sought safety by hiding in houses and tombs. The city’s
governor, accompanied by his garrison troops, came out to see what was
happening and put himself at the front of the crowd. Prisoners from the
local jail, who escaped when they heard the city was taken, joined them. The
augmented throng rushed to the east gate, where attack was most likely,
only to find that nothing had happened. From there they hastened to the
other gates, but found no Mongols. It finally turned out that the rumor had
started when people heard a group of women, in procession behind an
image of the Virgin, imploring God in tones of lamentation to spare the
people from the Turks and Mongols. Those who heard thought the women
were weeping because the city had actually been taken, and spread the
rumor that caused the panic. When the news of these events reached
Constantinople, the emperor castigated the population, pointing out how
irrational their fears were since the Mongols had only just arrived in Asia
Minor from Persia, warning them to stay on their guard in the future.
The incident reveals not only a state of mind—a population swift to panic at the
thought of attack from the east—but a present reality; the long-established
enemy, the Seljuks, with whom a modus vivendi had been reached, were now
weakened and largely replaced by the far more formidable Mongols at a
time when the Byzantine defenses were being undermined by the transfer
of men and resources to Europe. Although Nicaea was still at some distance
from the frontier, the defensive system no longer inspired confidence in the
face of Mongols and the bands of Turkish tribesmen released by the weak-
ening of central power in Anatolia.16
The Mongols had inspired special fear since 1256, when a Mongol army
invaded Asia Minor where they settled on new grazing lands, and when
the Mongol lands of Iran and Anatolia were organized into the Ilkhanid
sultanate. After a vain attempt to resist them, in which the self-exiled
Michael Palaeologus took part, the Seljuk sultan Izz al-Din fled to the rela-
tive safety of Byzantine territory. Two years later, Mongol power looked
even more overwhelming when they took Baghdad and brought an end to
the 500-year-old Abbasid caliphate. Yet the greatest danger for Byzantium
lay not in the manifest power of the Mongols, who represented a regime
that could be dealt with, but in the insidious infiltration of the lawless
Turcoman tribes.
For a century (with a few exceptions) relations between Byzantium and
the Seljuks of Rum had been stable, even favorable. That was particularly
true of the years from 1211 until the appearance of the Mongols on the
scene. The Byzantines rapidly understood the changed situation and began
to shift their alliance from Seljuks to Mongols. Already John Vatatzes
(1222–1254) had ordered fortresses to be well stocked with food and weap-
ons against the arrival of these unknown people, whom some thought had
dogs’ heads or were cannibals. More realistically, in 1257, Theodore Lascaris
exchanged embassies, in the process trying to impress the Mongols with the
power and splendor of Byzantium.17 This led to a treaty between Michael
Palaeologus and the Ilkhanid sultan Hulagu in 1260, which recognized the
Mongols’ dominion without even mentioning the Seljuks, who disappear
from Michael’s diplomacy.18 Diplomatic exchanges continued, culminating
in a marriage alliance in 1265 with the emperor’s illegitimate daughter
Maria sent as bride to Hulagu; when he died before her arrival, she was
married to his son and successor Abaqa.19 At the same time Michael had
also entered into relations with the Mamlukes of Egypt, anxious to secure
the trade route to the Black Sea, the source of the slaves who formed the
major element of their military and ruling class.20 Around 1270, Michael
sent an embassy with rich presents to Nogay, commander of the Golden

Horde north of the Black Sea, to secure protection against the king of the
Bulgars who was threatening imperial territory. This led to another marriage
alliance where another illegitimate daughter of the emperor was married to
what the Byzantines called a Tatar.21 By these means, Michael established
good relations with all the local great powers, whatever their mutual hostil-
ity may have been. More immediately, he protected his eastern frontier at a
time when he faced major threats from the west; he was safe from the
Mongols, but not from the tribes.
Michael recognized that much of the problem of the eastern frontier
stemmed from corruption (government officials pocketed the subsidies for
the frontiersmen) and falsely optimistic reports of the situation, for example
that lost places could easily be recaptured. In 1280 and the two following
years, he personally went to the Sangarius frontier, to inspect and bolster
the defenses.
In 1281 after a sally across the river that only drove the Turks into tempo-
rary retreat, he decided that pursuit into such rough and deserted country,
where he found only the camps of the nomads, would be fruitless. Instead,
he built forts on both sides of the river and blocked its banks with branches
of trees. He returned to Prusa. The next year, he strengthened the riverine
defenses, then withdrew to Lopadion in Mysia, where he planned to fortify
the region of Achyraous, but died soon after. This land was far west of the
frontier, an indication of the vulnerability of the imperial possessions.22
Meanwhile, in 1280, Michael sent out an army under his son Andronicus to
rescue the empire’s Asian possessions from a drastic situation: the Maeander
and Cayster regions with Antioch, Miletus, and Priene had already been
lost, as had Magedon, because of the paucity of defenders, while the districts
closer to the capital were seriously weakened.23 According to Marino Sanudo,
writing in the 1330s, Michael had left the Maeander region without protec-
tion against the frequent attacks of the Turks whose leader, “Turquenodomar
Mandachia” had taken control; he mentions especially the large lake
abounding in fish which generated substantial revenue for taxation.24 The
prince’s expedition started well: he defeated the Turks “between Ionia and
Lydia,” advanced to Laodicea, secured Philadelphia and cleared the Cayster
valley of the enemy. Then, however, he embarked on an extravagant project
to rebuild the city of Tralles, largest of the Maeander region. He squandered
his effort and resources on rebuilding and repopulating the city, but failed to
provide it with an adequate water supply; he then returned to Nymphaeum
and the capital, still in 1280. Consequently, when the forces of “Salpakis
Mantakhias” (i.e., Menteşe) arrived on the scene in 1284, after defeating the
Byzantines at Nysa, some thirty kilometers upstream, he had no difficulty in
capturing and devastating Tralles, which long remained abandoned.25 So
far, the accounts of the numerous expeditions to the frontiers describe their
adversaries as “Persians” (i.e., Turks) or “enemies”; Menteşe is the first
leader to be mentioned by name—and the only one until Osman and others
appear almost twenty years later.
Andronicus II, now emperor, returned to Bithynia in 1283/4, leading his
forces to Nicomedia, the Sangarius, and Nicaea. During his march from
there to Lampsacus and Adramyttium, he defeated Turks in “Lydia.”26
In 1290, Andronicus embarked on a long tour of his Asiatic provinces.
He stayed in Nicaea, visited the Sangarius frontier, then moved via
Lopadion to Nymphaeum, which he made his headquarters. Altogether,
he stayed three years, mostly in regions that were still relatively secure.27 It
was probably then that substantial repairs were made to the walls of
Nicaea28 (Fig. 2.1).
Fig. 2.1 Nicaea: tower of Andronicus II

To deal with the most threatened districts—notably in the upper

Maeander where the Turks were crossing in large numbers—Andronicus sent
his nephew Alexius Philanthropenus with full powers over the Maeander
and Hermus regions in 1293.29 This began auspiciously: Philanthropenus
gained the trust of the troops, whom he paid well. He drove off a Turkish
attack on Achyraous, then moved on to Philadelphia without trouble.
Between there and the Maeander, he won a major victory. In 1295 he took
Priene and the island fortress of Duo Bounoi (Fig. 2.2) the latter from the
widow of “Salampakis” who guarded her husband’s treasure there, as well as
the city of Miletus, many of whose Turkish defenders joined him out of fear
of the Mongols.30 He pacified and secured the Maeander and Cayster val-
leys, regaining territory that had been lost since 1280 or before. But when
his troops, sick of the incompetent and corrupt government, wanted him to
become emperor, he raised the standard of revolt. His movement soon col-
lapsed when Livadarios, governor of the Neocastra, pretending to join him,
seized and blinded the successful general. The expedition had lasted more
than two years.31 The Turks who had joined Philanthropenus, having lost
their goods and property, devastated the frontier regions so severely that
Pachymeres wondered whether this paved the way for the massive irruption
of Turkish tribes that followed in the next decade.32
Fig. 2.2 Menteşe’s island fortress

Philanthropenus had come at a good time for defeating or suppressing

Turks in the Maeander region for in October 1291 the Ilkhan Gaikhatu,
who had been governor of Anatolia, set out with a huge army to restore
order in his southwest frontier.33 He captured Laodicea and massacred its
population and devastated the lands of Menteşe among others. His campaign
lasted until June 1292. This ferocious expedition would have left Menteşe
and its neighbors seriously weakened. This makes it easy to understand the
Turks’ fear of the Mongols and their willingness to join Philanthropenus.
In 1296, Andronicus set out again for the east, this time heading for the
Black Sea port of Chele, but only got as far as the suburbs of the capital
when the whole region was struck by a devastating earthquake, a bad omen
that precluded further advance.34 Yet there was still hope: in 1298, a new
leader John Tarchaniotes took the field with a different policy. He found
that the big landholders had gotten rich but accomplished very little, while
the small peasants, essential for defense, were struggling. He therefore
carried out an equalization, redistributing land and wealth. He was having
some success rebuilding army and fleet when jealousy and intrigue brought
him down. Prominent among his enemies were the landowners and officials
who had profited from corruption—and continued to do so even more.35
Pachymeres’ narrative becomes especially detailed for the years 1302–1305,
when Byzantine power in Asia Minor finally collapsed, but is silent for the
five preceding years, when the Turks first broke through the frontiers on a
large scale. In 1298, when Tarchaniotes made Pyrgion his base, the empire
still controlled much of Ionia, Lydia, and Bithynia; five years later, it was in a
desperate situation, its remaining territories being attacked from all sides.
Plainly, something had happened, but Pachymeres is silent, only hinting
that the revolt of Philanthropenus lay in the background. For Gregoras,
looking back at these events, it seemed that the tribal leaders were operating
in collaboration against the empire.36 These years were a time of widespread
revolts against the Ilkhans, a situation from which the tribes could be
expected to profit as the central government was distracted or weakened.
The most dangerous among them was the revolt of the governor Sulemish
(November 1298–April 1299; summer–autumn 1299) who for a moment
gained control of all Anatolia, with particularly destabilizing effect since he
had the support of the Turcomans of the frontier—that is, the region
between the central Anatolian plateau and the sea.
There is evidence for at least one strategic area leading a normal existence
as late as 1301. An inscription of that date from Alexandria Troas (near the
mouth of the Dardanelles) records the sale of vineyards and fruit trees by a
local (unnamed) monastery to one Manual Gavras for forty hyperpera, a

sum which he subsequently donated back to the monastery for the salvation
of his soul.37
It is striking that a whole range of Turkish chiefs suddenly break through
the Byzantine frontiers at the same time. Perhaps Gregoras knew what he
was doing when he wrote of collaboration. Four of these chiefs—Aydın,
Menteşe, Sasa, and Germiyan—were all connected by blood, marriage, or
subordination. They could, in theory, have worked together, devastating
and occupying the Aegean region. But that leaves Osman and the
Paphlagonians. If the whole lot were to operate jointly, there was only one
force that could compel them. That is, the descendants of Genghis Khan,
the Ilkhanid Mongols, whose domain stretched from Byzantium to eastern
Iran. But in principle they were hostile to the tribes, who had joined the
revolt of Sulemish, and their own ambitions were directed toward Syria
rather than Asia Minor.
In 1302, a body of over 8000 Alan warriors, who had been employed by
the Mongol Golden Horde, offered their service to Andronicus. He received
them with enthusiasm; they seemed like a windfall at a time when Asia
Minor was in critical condition. The emperor sent the elite of the Alans,
along with Byzantine troops, with his son and co-emperor Michael IX, to
Magnesia in April. Here, they had to face harassment from the (unnamed)
Turks who could flee to their hilltop fortresses. The imperial forces hesitated
to attack, ostensibly afraid of endangering the life of the emperor. Eventually,
virtually under siege in Magnesia, they decided to withdraw, on which news
the Turks raided as far as the Menemen plain, halfway to Smyrna. Finally,
the Alans determined to leave, being used to rapid successful campaigns
with the Mongols rather than inactivity. This came at a time when
Amourios, Lamises, “Atman” (Osman, his first appearance in a historical
source), and many others were attacking imperial territory.38 Eventually, in
the winter of 1302/3, the emperor slipped away at night to the relative
safety of Pergamum. Of the panic-stricken crowds who tried to follow,
many succumbed to the winter cold, the stampede or were captured by the
Turks. After this, many left their homeland for Pergamum, Adramyttion
(on the sea) or crossed the Hellespont to the relative security of Europe,
abandoning their property and means of livelihood.39 For some, the exo-
dus had begun early in 1302 when the inhabitants of Pylopythia—the
region of Yalova on the sea of Marmara—took refuge in the Princes Isles
off Constantinople, only to be attacked and looted there by pirates allied
with Venice.40
Meanwhile, Leo Muzalon, governor of Mesothynia (the land between

Nicomedia and Constantinople), with a small force of Byzantines and Alans,
was passing through Nicaea enroute to Nicomedia. After being ambushed
twice on the way, he came face-to-face with a Turkish force much larger than
his own at Bapheus near Nicomedia on July 27, 1302.41 Osman was joined by
men from Paphlagonia and the Maeander, the two regions with large popu-
lations of nomad Turcomans independent of any central authority.42 The
battle was a Turkish triumph: although the Alans fought bravely, Muzalon and
the survivors rushed to take refuge within the walls of Nicomedia, leaving
Osman victorious for the first time over a Byzantine army.43 At that time, the
governor of Nicomedia was a converted Mongol, “Koutzimpaxis,” whose
daughter was married to “Solymampax,” the leader of a band occupying ter-
ritory around Nicomedia; they apparently did not play a role in this battle.44
In the general picture of loss and misery, it is easy to lose sight of the
obvious. The battle of Bapheus may have marked a step on the downward
course of Byzantine Bithynia, but Nicomedia, where Osman won his victory,
did not fall to the Turks until 1337, thirty-five years after Bapheus. Likewise,
the beleaguered cities of Prusa and Nicaea held out against all odds for
twenty or more years.
The situation in Bithynia, though, was certainly dire: defense of the civil-
ian population collapsed, the local economy was disrupted as houses and
crops were destroyed, and the rural population fled to wherever they could
hope to find security—to the fortified cities of Nicomedia, Nicaea, and Prusa,
or to Constantinople and the islands. Mysia, west of Bithynia, also offered
refuge in the inland fortresses of Lopadion and Achyraous, or in the sea-
ports of Cyzicus and Pegae. But the Turks ravaged the open countryside as
far as Adramyttion where the emperor had established his headquarters. In
all this, the Turks came away with immense quantities of prisoners, animals,
loot, and harvested crops. Yet they still hesitated to attack the peninsula
west of Nicomedia and avoided the suburbs of the capital.45
This after they had shocked the residents of Constantinople on December
13, 1302, when a force of Turks appeared at Scutari across the Bosphorus
from the capital. Their leader was probably Osman, active in the area, but
the raid had no aftermath.46
The sea was rapidly becoming less safe, for the Turks started to build
ships and in 1302 to use them to raid the offshore islands of Chios, Samos,
Carpathos, and Rhodes as well as the Cyclades which they devastated and
virtually depopulated.47 These may have been the actions of Menteşe, well
established on the coast of Caria.48
Pachymeres offered a gloomy prospect of the year 1303: devastation was

everywhere and every day got worse: apart from a few strongholds, Bithynia,
Mysia, Phrygia, Lydia, and Asia—all of Byzantine Asia Minor—were ruined.
The historian here enumerates the agents of this disaster, the heads of the
incipient Turkish states then being carved out of Byzantine territory,
describing them by the names of their leaders in the plural.49 Since they
will be met again, it is worth looking at them more closely in the order
Pachymeres presents them.
Amourioi: Ales Amourios and his Brother, Sons of Amourios

Atmanes: the followers of “Atman,” i.e., Osman
Atinai: Aydın
Alisurai: Alişir of Germiyan
Mantakhiai: Menteşe
Salampaxides: perhaps another name for Menteşe50
Alaides: unidentified, but Alais will appear below, ravaging the Hermus
valley
Ameramanai: probably Emir Yaman, founder of the Candaroğlu dynasty
of Paphlagonia.51
Lamises: unidentified; appeared in 1302 with Osman
Sfondulai: unidentified
Pagdinai: perhaps Karesi52 . . . and other cursed names.
The local landholding defense forces, who fled as their houses and lands
were devastated, could not be replaced in the prevailing chaos, nor was it
possible to come to terms with the invaders because there were too many
different bands and leaders, and even if a deal could be made, any tribes-
men who did not like it would decamp and join another band. Meanwhile
the co-emperor Michael IX had reached Cyzicus, whose archbishop had
organized local defenses and care of the flood of refugees. But Michael, in
fear of Turkish attack, left the city for the greater security of the heavily for-
tified seaport of Pegae, further west, where he fell ill.53 He only returned to
the capital in January 1304, having accomplished nothing.
The Aegean regions were not much better off than Bithynia. In 1303, a
chief named Alais was ravaging the Hermus valley when news arrived that
the emperor was making an alliance with Ghazan, khan of the Mongols.
Looking for a safe place for his men and the loot they had accumulated, he
came to Sardis, which had a powerfully and virtually inaccessible acropolis,

where the local population had taken refuge (Fig. 2.3). Alais proposed to
the defenders to share the castle. They reluctantly agreed since an arrange-
ment with the Turks would allow them to bring in water and sow their
crops, while the Turks promised to leave them in peace, attacking only oth-
ers. A wall was built in the citadel to keep the newcomers separate from the
natives, but when it appeared that the Mongol threat was fading, the Turks
planned to turn on their neighbors. The locals managed, however, to send
word to the imperial commander Nestongos Doukas, who attacked the for-
tress by night and disposed of the Turks.54
Alais was not alone in his fear of the Mongols. A Turkish chief from
Paphlagonia, known only by the Byzantine form of his name, Ales
Amourios, had put together a band of fighters and defeated the son of the
former Seljuk sultan Izz ed-Din, who himself had killed Amourios, father of
Ales Amourios, all this apparently in the region between Kastamonu and
the Sangarius in the last decade of the thirteenth century.55 In alliance
with the Turks of Kastamonu, he had devastated the regions beyond the
Fig. 2.3 Aerial view looking north toward Acropolis of Sardis. © Archaeological
Exploration of Sardis/President and Fellows of Harvard College, reproduced
with permission
Sangarius, but refrained from open hostility against the empire, whose
frontier along the Sangarius was in any case heavily fortified. In March
1302, however, unwonted spring floods caused the river momentarily to
change course and brought down masses of rocks and debris that made it
possible to cross and caused the defenders of the fortresses to flee.56 Ales
Amourios at first held back, but when he heard of Osman’s successes
between Nicaea and Nicomedia he abandoned his agreement with the
emperor and allowed his men to join in the general devastation. It was
probably this activity that caused him to be included among the chiefs who
were ravaging imperial territory. Two years later, though, when news
arrived of the death of Ghazan, he requested the district of Mesonesion west
of the Sangarius from the emperor as a place where his forces could settle in
security and provide protection for the imperial lands beyond. He did this,
according to the historian, because he wanted to ingratiate himself with
Byzantium or in fear of what a new Mongol ruler might bring. While wait-
ing for a reply, some of his men moved in anyway, harassing the local popu-
lation who were trying to harvest their crops.57
Alais and Ales Amourios had reason to worry, for Andronicus was fol-
lowing his father Michael’s policy of alliance with the all-powerful Mongols
against the Seljuks and Turcomans. Michael VIII had already betrothed his
illegitimate daughter Maria to the Ilkhan Hulagu. But he died in 1265 before
the princess reached his court. Instead, she married Hulagu’s son Abaqa,
and stayed with him until his death in 1282. Subsequently, she returned to
Constantinople where she founded a monastery; she will reappear in this
narrative.58 Andronicus, faced with the imminent ruin of his position in
Asia Minor, with the immediate threat to Philadelphia, and despairing of
his armies and mercenaries, proposed an alliance with Ghazan, Ilkhanid
ruler since 1295, who had recently suppressed a widespread revolt in
Anatolia. The emperor’s natural daughter was to marry Ghazan who in turn
would help against the Turks. But before the negotiations were complete,
Ghazan suddenly died at the age of thirty-two in May 1304. This was devas-
tating news, until the court learned that Ghazan’s brother and successor
Uljaytu Khodabende (whom the Greeks called by his personal name
Kharbanda) planned to carry on his brother’s policies. Consequently,
another embassy was sent proposing the same marriage alliance and asking
for armed assistance, with results that will appear next.59
By that time—and destined to make things even worse for Byzantium—a
heavily armed professional force, the Catalan Grand Company, engaged by
the emperor in a desperate effort to restore his failing fortunes in Anatolia,
had landed in Cyzicus in September 1303, under the command of their

Grand Duke Roger de Flor.60 There, they found a broad, fertile, and well-
populated peninsula highly desired by Turks who were camping nearby.
The Catalans struck quickly and mercilessly: they killed the men and sent
the women and children as slaves to the emperor. In the spring of 1304, they
set out for Achyraous, then south over the mountains to Germe where they
hoped to trap the Turkish force that had occupied the town, but it escaped.61
From there, they marched through Chliara (Fig. 2.4) to Philadelphia where
the local fortresses had surrendered to Alişir of Germiyan, most powerful of
the Turkish chiefs.62 Among the losses was the stronghold of Tripolis which
commanded the pass from the Maeander to Philadelphia63 (Fig. 2.5).
The city was well fortified and had laid in abundant supplies, but these
were not sufficient when subjected to a long siege. Eventually the locals
came to terms with the Turks, allowing their merchants to enter the city.
Alishir, however, took advantage of the situation to smuggle in weapons and
to take Tripolis, which he made his base.64 In a battle at Aulax near
Philadelphia, the Catalans were victorious, forcing Alişir, wounded, to with-
draw to Tripolis, where he organized harassing attacks before retiring to
Amorium in Phrygia. The Catalan chronicler of these events reports that
Philadelphia (whose walls were eighteen miles long!) was under attack by
Fig. 2.4 Chliara: a Byzantine military base, secure in the mountains

overlooking the plain [40]
Fig. 2.5 Tripolis
the bands of “Sasa and Tin”—i.e., Sasan, who will be met again shortly, and
Aydın65 (Fig. 2.6). This does not contradict the narrative of Pachymeres,
which features Alişir, for both Sasan and Aydın were allies or subordinates
of Germiyan.66
The Catalans did not pursue Alişir out of fear of ambushes. Instead, they
spent two weeks in Philadelphia where they were welcomed, moved west to
the region of Kula, then to Nymphaion and Magnesia, and on to the Cayster
valley towns of Pyrgion and Thyraia, where they beat off a Turkish attack by
men who had escaped from the battle at Philadelphia joined by tribesmen
of Menteşe, and to Ephesus. Everywhere the Catalans went they were
welcomed by the people, whom they proceeded to treat savagely, demanding
as much money as they could forcibly extract. They were especially harsh to
the commanders of cities who had by necessity been forced to yield to
the Turks.
When he was in Thyraia, the Grand Duke received news that Catalan
reinforcements were on their way. He went down to the port of Anaia to
meet them and there won more victories over the Turks.
Like the leaders of the marauding Turkish tribes, Roger needed a safe
place for his loot and supplies. He chose Magnesia whose long walls enclosed
a city in a naturally strong position at the foot of a mountain (Fig. 2.7). But
when he returned there after a successful expedition, he found the city in
Fig. 2.6 The walls of Philadelphia. © History and Art Collection/Alamy

Stock Photo
Fig. 2.7 Walls and citadel of Magnesia © History and Art Collection/Alamy
Stock Photo
the hands of a certain Attaliotes, an imperial equerry who was in revolt against
the emperor and had prevented the governor Nestongos Doukas from
entering the city.67 He soon came to terms with Roger, to whom be submit-
ted. The cause and the circumstances of the city’s revolt are unknown, but it
seems probable that the locals were taking defense into their own hands as
the emperor proved unable to save the country from the devastation
inflicted by the Turks. It is unlikely that Magnesia was alone in breaking
with Constantinople. If many did, it would offer some explanation for the
Turks’ mention of tekfurs as if they were independent commanders of
their cities.
In the Summer of 1304, after conquering and looting Pyrgion and
Ephesus, Roger returned to Magnesia where he had left horses and a large
treasure, only to find the gates shut against him, for the Catalan reputation
for ferocity and extortion had preceded him.68 The locals, trusting in a
detachment of Alans, a year’s supply of wheat and a secure source of water,
and bolstered by the prospect of seizing Roger’s goods, prepared for a siege.
Attaliotes urged them on. They massacred all the Catalans in the city. Roger
attacked with siege machinery; the Magnesians responded with the jeers
and mockery he detested. As the siege dragged on, the Turks returned, dev-
astating the countryside and leaving only a few terrified populations behind
their walls.
At this point, a message arrived from the emperor, requesting the
Catalans to come to his aid in a war with Bulgaria. They agreed and set off
by sea to the Dardanelles, leaving strong garrisons in the cities they had
conquered in Anatolia, and planning to return there the following spring
(1305). Roger abandoned the siege of Magnesia, extorted money from the
cities on his route, crossed to Mitylene and then to the Gallipoli peninsula,
where the Catalans caused even greater trouble for Byzantium.69 Meanwhile,
a body of Alans wound up at Pegae, where they camped outside the walls
and managed to defeat the attack of a much larger Turkish force.70 In
Magnesia, Attaliotes was still in charge, revolting against the emperor, after
the departure of the Catalans; his ultimate fate is unknown.71
The Catalan victories proved ephemeral as the provinces they rescued
suffered almost as much from their exactions as from Turkish attack, left
less capable of resisting attack, even if they wanted to; for them, imperial
rule was a mixed blessing indeed.
Two isolated incidents at this time give a rare view of conditions in the
far northwest of Anatolia. A certain Machrames, otherwise unknown but
described as an important servant of the emperor, lived on the Scamander
river in the Troad. When the Turks, who had occupied the region of Mount
Ida, advanced, he took refuge in Assos, where he was welcomed and put in
charge of defending the besieged coastal city. Eventually, the despairing
defenders moved off to the apparent security of nearby Mitylene and
Machrames reluctantly joined them. But by supremely bad luck, the Catalans
were there. They seized Machrames, demanded a huge ransom, and, when
he could not pay, tortured and executed him, Assos, left empty, was presum-
ably taken by the Turks who, ironically, appear to have named it Behramkale
after its unfortunate defender.72
Still in 1304, an adventurer named Choiroboskos put together a small pri-
vate army of men armed with bows and clubs believing that he could pick off
isolated detachments of Turks. He approached the fortress of Kenchreai, where
much of the population of the Scamander valley had taken refuge. The defend-
ers welcomed the new force, which at first drove off the Turks, but when the
Turkish horsemen came back in larger numbers, Choiroboskos was captured
and killed because he couldn’t raise the demanded ransom. The Turks captured
the town by cutting it off from its water supply, massacred its defenders, looted
the place, and burned it down.73
Pachymeres surveyed the losses of 1304 in a pessimistic account rich in
toponyms.74 No place beyond the Bosporus was safe as Turks camped where
they liked, attacking in small groups, hard to catch, rather than mass cam-
paigns where they might be met in the field. Chele and Astrabete, on the
Black Sea, and even Hieron at the entrance to the Bosporus were under
attack. Belokome, Angelokome, Anagourdys, Platanea, Melangeia, and all
places around were emptied of inhabitants, while Kroulla and Katoikia suf-
fered as much or worse. This panorama of devastation evidently represents
the situation in Bithynia, tantalizingly naming places of some importance in
these struggles. The seaports pose no problem; Melangeia (known also as
Malagina) is on the Sangarius; Kroulla lies just southwest of the lake of
Nicaea; but the rest remain unidentified. But note that Angelokome is not
Inegol, nor is Belokome Bilecik; Katoikia is apparently not Kite.75
Nicaea was increasingly isolated: the roads from Neakome and Heraklion
on the Gulf of Nicomedia were closed; Pylopythia (the district of the major
port of Pylae/Yalova) suffered like the region of Nicomedia. The only route
available was the overgrown and disused road from the port of Cius, where
travelers spent the day, then proceeded by night to Nicaea, over the lake to
the only gate that was safe to open. The emperor sent Sgouros, a commander
of crossbowmen, to relieve Katoikia raising the hopes of its defenders, but
the enemy arriving in force blocked the roads and crushed Sgouros’ force
whose survivors fled the scene. Osman, returning from this battle, attacked
Belokome which had taken the imperial side, killed the defenders and took
control of the fort and a large treasure, gaining security from the fortresses
he now controlled. While these small places were abandoned or conquered,
widely separated cities like Prusa and Pegae suffered from the influx of
refugees, exacerbated in Pegae by plague and famine.76
In 1305, Kouboukleia, northwest of Bursa and long since fortified, was
attacked by the Turks of Atares, an unidentified leader. The town’s only hope
was from Lopadion, where the governor had arrested half the Catalans left
when their commander Roger de Flor had hopes of carving out his own
domain in Anatolia. He sent the other half against the Turks, but they
deserted, joined the attack, took the town, and killed its defenders.77
In one case at least, the locals took defense into their own hands. Late in
1305, a monk called Hilarion, who had been sent to look after the needs of
the monastery of Elegmoi (on the coast due north of Bursa) by its mother
church in Constantinople, found Turks looting the district daily. He organized
a local force, beat back the Turks, and took up watch over the countryside.
In doing so, he was violating a long-standing prohibition of monks taking
up arms and was forbidden to continue by both his abbot and the patriarch.
Hilarion appealed to the emperor who was sympathetic, but the process
took so long that the Turks returned in force, massacring all they could lay
their hands on. When he finally got permission to proceed, Hilarion forti-
fied the district as much as possible, but the Turks (evidently of Osman
though he is not named) occupied the whole surrounding district as they
attacked Bursa which was forced to pay a large bribe in exchange for what
the historian called the shadow of peace, not a real peace.78
A letter of the patriarch Athanasius I, written around 1306, confirms the
desperate situation of this coastal region. Reproving the metropolitan bishop
of Apamea for not going to his see, the patriarch notes that the entire region
was under attack by “wild beasts” and “Arabs” (i.e., Turks) and that amid
much bloodshed, the locals had taken refuge in the fortresses of Myrsine,
Syke, Rhodophyllon, and Muntania.79
Meanwhile, the emperor ordered the rich monasteries of the capital to
send their surplus grain to the starving population of Asia Minor. He was
also taking measures to renew an alliance with the Mongols. Their khan
Kharbanda (Uljaytu) responded favorably by putting together a force of
40,000 of which half, under his cousin, had reached Konya. He only waited
for the emperor to tell him where and against whom the entire force should
be sent.80 An alliance was all the more desirable because news arrived that
Sasan, who had asserted his independence from his father-in-law, Menteşe,
took Ephesus on October 24, 1305, after blockading and taking Thyraia in
the Cayster valley, to which he deported much of the population of
Ephesus.81 By this time, Turks had taken the entire coastland except for the
ports of Adramyttion and Phocaea, occupied by the Genoese Manuel
Zaccaria and guarded, Pachymeres reports, by the martial bravery of the
Italians.82 The situation in Bithynia was no less dangerous. Andronicus sent
a relative by marriage, Kassianos, to regulate the situation of Mesothynia,
but this proved abortive when the general, denounced by a tax collector he
had maltreated, took refuge in Chele where he was arrested and brought to
the capital.83
During the next two years, the Mongols became an important part of the
scene. In 1306, as the emperor was preparing a marriage alliance with them,
he learned that Nicaea was under threat. He sent an army along with his sis-
ter Maria, (known as the empress of the Mongols because she was widow of
the previous Ilkhan) to make arrangements for the marriage of the princess
with Kharbanda, and to deal with the Turks. Once installed in the city, she
threatened to call in the Mongols against Osman.84 When he learned that a
Mongol army had indeed moved into Asia Minor, Osman attacked fero-
ciously, tearing up vines and destroying the harvest. He moved against
strategic Trikokkia, the “rampart of Nicaea,” and took the place by filling
the ditches in which the defenders had confidence, in the summer of 1307.
He slaughtered the defenders and, filled with self-confidence, he awaited the
Mongols, if they should come.85 By then, he had conquered the entire region
between Nicaea and the sea.86 The 30,000 Mongols sent by Kharbanda now
arrived on the scene, with a greater effect than hoped: they pushed back the
Ottomans who were forced to abandon all the Byzantine forts they had occu-
pied and to take refuge in the Bithynian Olympus above Prusa.87 This was
the first defeat Osman had suffered; it held him back for the next twenty years.
At this point, the story takes leave of its primary source, Pachymeres. He
leaves a scene where Prusa is paying tribute for a dubious security, Nicaea
temporarily rescued from blockade, and the Ottomans forced to withdraw
to the mountains. In 1307, imperial fortunes in Bithynia are on the verge of
collapse, but . . . nothing happened. When the far less informative Gregoras
carries on the narrative, the situation has not changed: in 1326, when
Ottomans reappear on the scene, Prusa, Nicaea and Nicomedia are still
under threat; Bithynia is somehow holding out, though elsewhere Byzantine
power is fading rapidly. This hiatus will need explaining, especially since the
Ottoman sources are equally uninformative about these years.
There appears to be only one sole mention of Bithynia in these intervening

years, a decree of the patriarch and his synod of October 1318 directed to
the metropolitan bishop of Prousa (Bursa).88 The authorities, recognizing
the irregular and straitened circumstances of the time, assign the neighboring
bishopric of Apamea (Mudanya on the Sea of Marmora) to the jurisdiction
of Prusa, requesting the bishop to look after Apamea’s Christian population
and monks. In other words, the bishopric of Apamea could no longer be
maintained independently but needed to be put under the care of its larger
neighbor. This indicates that Prusa had a functioning church, was still under
Byzantine control, and had access to the sea. It seems not to have been
under blockade. Nothing more is known.
Bithynia may have been quiescent, but there was no lack of activity else-
where as the Byzantine position in Anatolia slid toward total loss. The
Catalans had been a last hope, but when they left, the Turks renewed their
attacks. Already in 1305, Ionia, with Ephesus, Pyrgion, and Thyraia, was in
Turkish hands, as were the coastal regions. The remaining imperial outposts
were in a desperate situation, despairing of help from Constantinople which
had no armies to send. By 1308, when many fortresses had been under attack
for four years, they called on Charles of Valois, the pretender to the imperial
throne (through his relationship to the defunct Latin Empire). When he also
was in no position to help, the Hermus valley, with Magnesia, Sardis, and
Nymhaeum fell to the Turks by 1310.89 According to the traditional interpre-
tation of the sources, however, Magnesi was taken in 1313, to be followed
two years later by Nymphaeum, where the Lascarids had built their palace.90
In 1310, Philadelphia, saved and plundered by the Catalans six years
before, was under siege again, suffering from a growing shortage of supplies.
The defense was led by the metropolitan bishop Theoleptos who personally
went to the Turkish commander. Out of respect for the prelate, he agreed to
lift the siege and to accept a tribute instead. The money was used to build a
medrese in Germiyan’s capital, Kütahya.91 Germiyan and Aydın returned in
1322, when they besieged the city and its outlying bulwark, Fort St. Nicholas.
Although they employed catapults and made the inhabitants miserable for a
year and seven months, they failed to conquer, for the emperor, short of
men and money and unable to send a force so far into the interior through
enemy territory, sent the aged Philanthropenus who had such a favorable
reputation among the Turks that they abandoned the siege and Philadelphia
was able to prosper. It drew its wealth from its fertile territory and success-
ful trade in high-quality textiles, especially silk, and leather. The siege had
lasted a year and seven months.92
Soon after, in 1327, the ferocious Mongol governor of Anatolia, Timurtash,

arrived in Philadelphia, not to attack the city but to bring Germiyan and
Aydın to order and to support the independence of Philadelphia, subject to
the emperor who was allied with the Mongols.93 Mongol influence, however,
was nearing the end, for the Ilkhanid state fell apart after the death of its last
effective ruler, Abu Sa’id, in 1335.
The momentous fall of Prusa received only scant notice when it suc-
cumbed on April 6, 1326: “In these days when no one was taking care of the
East, most of the cities and lands of Bithynia fell to the Turks. Prusa, a city
besieged by hunger, was taken.”94 Likewise, the fall of Lopadion, which con-
trolled the strategic bridge over the Macestus that had marked the western
limit of Osman’s conquests, in April 1327, is only reported by a sentence in a
short chronicle which does not even tell to whom it fell (most probably
Orhan but conceivably Karesi).95 If “no one was taking care of the east,” it
was because the empire was embroiled in a civil war between the old
emperor Andronicus II and his grandson Andronicus III. It began in 1320
and was only resolved in 1328.
In 1328, the new emperor Andronicus III sailed to Cyzicus where he
inspected a region he had not seen before, but more importantly intended
to enter negotiations with Temirhan, son of Yahşı, the emir of Karesi,
described as the ruler of Phrygia, who had been putting pressure on the
eastern cities of the Hellespont subject to the empire. He proceeded to Pegai
where he met the emir. Temirhan and his men dismounted from their
horses and prostrated themselves before the emperor. When the two rulers
met the next day and exchanged gifts, Temirhan promised not to attack the
Byzantine lands, a promise he faithfully kept.96
In the spring of 1329, Andronicus, frustrated by the constant Turkish
attacks on his territory and anxious to relieve Nicaea, then under direct
threat, decided to make a campaign into Bithynia. He sought the advice of
Kontophre, governor of Mesothynia, who, knowing the Turks extremely
well, told the emperor to move quickly before they made their annual
migration into higher ground, where they would be impossible to pursue.
Consequently, an army was recruited from the European provinces, with
2000 elite troops and a mass of poorly trained artisans and peasants. The
emperor in person led them, together with his right-hand man, the grand
domestic John Cantacuzene (the future emperor John VI whose eyewitness
account preserves details of the campaign). They marched for two days into
Bithynia, where they met the larger Turkish force in the plain of Pelekanon
on June 10. Orhan, forewarned of their approach, occupied the higher
ground. The Byzantines were successful at first but could not follow up their
advantage because of the mobility of the Turks and the configuration of the
land, full of hills and ravines suitable for ambushes. They celebrated a vic-
tory and planned to withdraw, but the emperor had an accident and rumor
flew that he had been fatally wounded. At that, the troops fled to the neigh-
boring fortresses of Philokrene, Niketiaton (Fig. 2.8), Dakibyza, and Ritzion.
They then rallied and attacked Orhan, who withdrew to his camp. The next
day, they returned to the capital. At first sight, this looks like a Byzantine
victory or at worst a draw, but in fact the expedition failed in its objective.
However superior the Byzantine army may have been, it could not inflict a
real defeat on the mobile enemy or establish a secure position outside the
fortified places.97
This was the last imperial offensive in Asia Minor. As soon as the army
was gone, the Turks returned to harass the country, block the routes, and
achieve their goal: Nicaea fell two years later, on March 2, 1331, after a long
siege that threatened the defenders with starvation. The Turks sold sacred
books and icons as well as the relics of two female saints. They extended their
control over the coastal regions and imposed heavy taxes on the towns.98
In 1330, Andronicus learned that Orhan had surrounded Nicomedia and
set up siege machines for a major assault. He promptly embarked foot sol-
diers and cavalry on whatever merchant ships he could find and set sail.
Fig. 2.8 Niketiaton

Defending the city was crucial for it played major role in the food supply of
the capital.99 As Cantacuzene approached Nicomedia, a message arrived
from Orhan seeking peace. Gifts were exchanged and an agreement made
that Orhan would be a friend of the empire and not attack its possessions.
Andronicus then visited the city which he had never seen, brought in
supplies and stayed a week. It was probably on this occasion that he agreed
to pay Orhan 12,000 hyperpera (Byzantine gold coins) a year to guarantee
security for the fortresses of Mesothynia, from Nicomedia to the capital.100
Three years later, as he was about to march against Bulgaria, Andronicus
received news that Orhan was preparing to attack. “Nicomedia,” wrote
Cantacuzene, “could not be taken by weapons or force because of its circuit
of extremely strong walls and the powerful nature of the site. It feared only
lack of provisions. The barbarians, who understood this, ignored the walls,
which they could not capture by siege, and hastened to occupy the
approaches by which the city was fed” (Figs. 2.9, 2.10). The emperor aban-
doned his Bulgarian plans; loaded men, horses, and grain on battleships
and freighters; and set out. The Turks withdrew on news of his approach.
Andronicus spent two days in the city, unloaded the grain, and encouraged
the defenders by his speeches.101
While this was going on, the empire faced increased problems from
Aydın. In 1317, Mehmet of Aydın had taken the hilltop fortress and city of
Fig. 2.9 Nicomedia: Byzantine walls

Fig. 2.10 The gulf of Nicomedia
Smyrna from the Byzantines by surprise.102 This did not involve control of
the port however, for it had a separate fortification under Genoese control
that Mehmet’s son Umur only captured in 1329 after a long siege. With that,
he embarked on a spectacular career that would involve building fleets,
attacking imperial possessions in Thrace, Greece, and the islands, and even-
tually entering into an alliance with Cantacuzene.103 In 1329 or 1330, his
brother Hızır supported Umur in a successful attack on Chios. The constant
brigandage was such a threat to the Christian powers that they formed a
union (Sancta Unio) to combat the Turks: after long negotiations, Venice,
the knights of Rhodes, Cyprus, and the king of France joined under the
patronage of the pope, put together a fleet and inflicted heavy losses on the
Turks, but when the pope died in December 1334, the union broke up and
Umur and his allies resumed their unwelcome activities.104
1335, when the Byzantine position in Bithynia was on the verge of col-
lapse, was a momentous year for relations between Umur of Izmir and the
empire.105 Early in the year, Umur, together with the son of the emir of
Saruhan, led a massive attack of 276 ships on Greece. In the summer he
attacked Philadelphia—a long siege which finally terminated in the city
agreeing to pay tribute. Later in the year, alarmed by the Turks’ success,
Cantacuzene met with Umur at Clazomenae, west of Izmir. He persuaded
Umur to renounce the tribute of Philadelphia and entered into an alliance
that was to prove permanent.106 Shortly after, the emperor Andronicus
himself had an interview with Umur in a galley off the coast of Kara Burun,
the mountain massif between Izmir and Chios. They confirmed the alliance,
directed at least in part against the Genoese who occupied Phocaea and
Lesbos.107
At this time, the emperor, with substantial support from Saruhan “who
ruled the lands east of Phocaea” had been besieging Phocaea (then in the
hands of the Genoese) during which he received a friendly visit from Umur,
Hızır, and Suleyman, the sons of Aydın. This led to the interview at
Clazomenae, after which the Genoese accepted terms: Lesbos and Phocaea
would return to imperial control, Saruhan would continue to provide supplies
to Phocaea, where the Genoese would retain their ancient privileges, along
with the right to trade freely throughout imperial territories.108 After the
agreement with Cantacuzene and the emperor, Umur received a subsidy, sub-
sequently provided mercenary troops, and from now on attacked Frankish
territories and even Bulgaria, but not the remaining Byzantine lands and
outposts. He was rapidly becoming a major power in the Aegean and Balkans.
At the beginning of summer 1337, as Gregoras recounts, the people of
Thrace had to suffer yet again at the hands of the Asiatic Turks. Those of
Ionia held back because of the treaty recently made at Phocaea, but the
Turks of Troy and the Hellespont (i.e., Karesi) loaded their men and horses
and attacked Thrace, from which they withdrew on terms after being beaten
in a skirmish. But the same summer brought terrible news from Asia: that
“Orhan son of Atuman, ruler of Bithynia,” had been secretly putting
together a force that would strike from Asia in two divisions: from Hieron
and from the Propontis, both directed against the suburbs of Constantinople,
which they had not reached before and there to loot and burn the harvests.
They planned to capture two fortresses close to the capital and use them as
bases for future attacks. The grand domestic Cantacuzene put together a
small force and with his few ships met the Turks at Rhegion, only 110 stades
from the capital, where he inflicted serious losses on them with only minor
casualties on his side.109
Gregoras rather casually tells of a major loss in Asia Minor, which marked
the collapse of the Byzantine position in Bithynia and definitive triumph of
Orhan: “At this time (1338), when the emperor was paying no attention,
Nicomedia, the metropolis of Bithynia, was taken, suffering from great hun-
ger because of the prolonged siege by the enemy.”110
In the summer of 1341, Cantacuzene learned that Saruhan of Lydia and
Yahşı of Pergamum were planning a joint descent on Thrace. To protect
himself from attack from behind while dealing with this threat, he sent an
embassy with a huge bribe to Orhan, “satrap of Bithynia,” who agreed to
make peace. Cantacuzene then met Yahşı’s troops who arrived in Thrace in
two waves, beating them so badly that the emir agreed to peace terms.
Cantacuzene’s fleet successfully dealt with Saruhan and ravaged his coastal
towns.111 The same year, Cantacuzene entered into alliance with “Aliseres
satrap of Cotyaeum” (Alişir of Germiyan) who promised to send infantry
and cavalry against Saruhan in an envisioned expedition to destroy their
hostile fleet.112 In all this, the emperor was using the traditional Byzantine
strategy of setting one barbarian against another, and finding that his bribes
could produce at least temporary alliances.
In 1345, during a destructive civil war that pitted Cantacuzene against
the dowager empress Anna, a certain John Vatatzes, who had been an ally of
Cantacuzene, switched sides, joined the empress, and attacked towns and
villages in Thrace loyal to Cantacuzene. He brought a large army thanks to
his close alliance with Suleyman, “satrap of Troy” who had married his
daughter.113 Unfortunately for him, Cantacuzene withdrew men and animals
into fortified positions, leaving the Turks little to loot. Feeling themselves
misled, they killed Vatatzes and joined Cantacuzene, once again illustrating
the unreliability of mercenary forces.114
By 1346, it was obvious that Orhan was the dominant figure in western
Asia Minor and in the best position to hurt or help Byzantium. That sum-
mer he sent an embassy to Cantacuzene, asking for his daughter in marriage
and proposing an alliance. Cantacuzene hesitated, asking advice. He sent to
his friend Umur for his opinion, who approved on the grounds that Orhan
could aid the empire easily and directly while he had to send his forces
across foreign territory (i.e., Saruhan). The marriage was celebrated at
Selymbria on the Sea of Marmora, and the alliance concluded. News of this
stirred the empress Anna, engaged in hostilities with Cantacuzene, so she
sent to Saruhan for a force to use against him. Saruhan sent the men, but
they wound up joining the other side, thanks to a stratagem employed by
Umur, who had pretended to join them. A casual mention in Cantacuzene’s
narrative shows that Heraclea and Amastris on the Black Sea were still in
Byzantine hands.115
In these years, Umur’s fleets constantly ravaged Western- controlled
Thrace and Greece, sometimes in alliance with the neighboring emirs of
Saruhan and Karesi.116 His exploits stirred such anxiety in the west that the
pope proclaimed a crusade against him in 1344. Although the Christians
failed to eliminate their nemesis, they did manage to capture the harbor fort
of Izmir, severely crippling his activities. This did not stop him from attack-
ing Philadelphia yet again, in March 1348, aided by his brothers. This time,
the Turks gained a foothold on the walls, but they were pushed back and
both sides came to an agreement by which Umur promised to leave the city
in peace. In fact, he actually planned another attack to take place over
Easter, but in the meantime, he left with the elite of his forces for Smyrna,
where he hoped to recapture the fort.117 During the battle, he took off his
helmet and received a fatal arrow. With that, the Europeans lost an enemy,
the empire a friend, and the Muslims their leading fighter.
As Nicephorus Gregoras wrote, “Umur was the most powerful of all the
satraps, exceeding the others in enthusiasm and daring. Ruler of Lydia and
Ionia, he filled the sea with his fleet, and in a short time his command of the
sea made him fearful to the Aegean islands but also to the Euboeans, the
Peloponnesians, the Cretans and Rhodians and of the whole shore from
Thessaly to Byzantium. He could raid them with his fleet whenever he liked
and extracted heavy yearly taxes from them. Soon, for Cantacuzene, whose
fame spread, with applause and long hymns, to land and sea, he was an
ardent and passionate friend. He promised to maintain his voluntary
friendship for his whole life to him and the children who succeeded him.
He kept his word to the end, in a way, I think, no other age could show.”118
This narrative ends on a pathetic note: late in 1352, the beleaguered citi-
zens of Philadelphia managed to get an embassy through to the pope, then
in Avignon, pleading for help against the Turks and offering to submit
themselves and their city to the pope and the Roman church in all temporal
matters in perpetuity. The pope unhelpfully replied early in 1353 that they
should abandon the schism (i.e., Orthodoxy) and recognize the primacy of
the Roman church in order to avoid the eternal punishments that were far
more serious than the danger from the Turks.119
Philadelphia managed to survive, but the next year, on March 2, 1354, a
devastating earthquake brought down the walls of Gallipoli and a force
commanded by Orhan’s son Suleiman, operating in the vicinity as ostensi-
ble allies of Cantacuzene, moved in and stayed. This occupation of a
European foothold turned out to be the beginning of a new age for Orhan,
Byzantium, Europe, and the whole western world.120
Notes
1. By far the most comprehensive and critical account of these events is that of
Arnakis 1947, 71–197.
2. Hunger 1976, 447–53; Neville 2018, 237–42.
3. Hunger 1976, 453–65; Neville 2018, 243–8.
4. Hunger 1976, 465–76; Neville 2018, 266–72.

5. The headquarters of this district were Adramyttion on the coast and Pergamum
and Chliara in the Caicus valley. For some of its fortresses see Foss 1982, 166ff.,
185–8 and Foss 1998, 160–6.
6. See Cahen 1951 and Korobeinikov 2014, 220–7 et passim.
7. Acropolites 315.
8. Pachymeres I.3–5 (1.27–35), III.22 (1.291–3). Reference to Pachymeres will
consist of two numbers: the first gives the traditional book and chapter; the
second denotes the volume and page number of the Failler edition.
9. Pachymeres II.6 (1.139f.).
10. Pachymeres III.12, 13 (1.259–67). Trikokkia is to be identified with Karatekin,
which occupies a strategic position on the only approach to Nicaea that could
be described as a mountain pass (zygos).
11. Between 1260 and 1295 there were at least seven Byzantine campaigns to the
Cayster/Maeander region, but only four to Bithynia. Most instructive are the
operations of 1304 when Andronicus II sent his son Michael IX to Magnesia
with a large force, including the majority of the Alan mercenaries, while
Muzalon in Bithynia got only a small force with fewer Alans.
12. Korobeinikov 2014, 226f.
13. Pachymeres III.21 (1.289–91).
14. Pachymeres III.22 (1.291f.).
15. Pachymeres IV.27 (2.403–7). For Abala, a fortress occupying a strategic location
on the road from the Hermus valley to Phrygia, see Foss 1979, 302–4; Magedon,
evidently in northern Lydia, has not been located. The Black Sea ports men-
tioned were Kromna, Amastris, Tios, and Heraclea. See also the more general
account of Gregoras I.138–42 who attributes the outburst of Turkish attacks to
the collapse of Seljuk authority at the hands of the Mongols and stresses
Paphlagonia as the base of the tribal forces.
16. Pachymeres III.28 (1.317–25).
17. Pachymeres II.25 (1.187f.).
18. Pachymeres II.24 (1.185). For relations between Byzantium and the Mongols
1260–1310, see Korobeinikov 2014, 203–16.
19. Pachymeres III.3 (1.235).
20. Pachymeres III.3–5 (1.237–43); Dölger Regesten 1902–1904 (Nov. 1261–
Nov. 1262).
21. Pachymeres V.3 (2.443–9); Dölger Regesten 1977.
22. Pachymeres VI.22, 25, 29, 34 (2.599, 623, 633–7, 657); Korobeinikov 2014,
248F. For the important fortress of Achyraous, see Foss 1982, 161–6.
23. I take Pachymeres’ endotero “more inward” as denoting territories in their rela-
tion to the center, Constantinople.
24. Hopf 1873, 145; see Wittek 1934, 47f., for interpretation of the name as
“Turkmen of the Sea.” For the lake, see n. 30.
25. Pachymeres VI.20–1 (2.591–9); cf. Korobeinikov 2014, 249f. For Salpakis, see
Failler 1994, 86f., and for the fate of Tralles, Lemerle 1957, 37.
26. See Korobeinikov 2014, 257–61 for the chronological problems associated with
this and the following expedition.
27. See Failler 1990, 15–28, with full references.
28. Foss & Winfield 1986, 112–14.
29. Korobeinikov 2014, 265–9.
30. “Salampakis” is apparently the same as “Salpakis” who took Tralles in 1284: see
n. 25. Planudes calls him “Salamates.” Duo Bounoi is to be identified with
Ikizler, the fortified islands in the Milesian Lake (formerly the Gulf of Miletus,
now Bafa Gölü): Wendel 1940, 438–43. See the description, with plans and
illustrations, of Wiegand 1913, 30–41.
31. Pachymeres IX.9–14 (3.237–57), Gregoras I.195–202; cf. Schreiner 1969,
376–83. Details of the expedition are given in the letters of the monk Maximos
Planoudes: see Korobeinikov 2014, 264–6.
32. Pachymeres IX.14 (3.257).
33. For these events in a different context, see “The Overlords,” pp. 157–162.
34. Pachymeres IX.15 (3.259f.).
35. Pachymeres IX.25 (3.285–9); Korobeinikov 2014, 269–71. Tarchaniotes made
his headquarters in Pygion, a powerful site overlooking the Cayster valley. He
eventually took refuge from his enemies in the local monastery of S. George
formerly the temple of Zeus, as Pachymeres writes, reflecting the ancient name
of the city, Dios Hieron.
36. Gregoras I.214: “the satraps of the Turks made an alliance (synaspismon),” “the
Turks, who had already made an agreement, distributed by lot all the land of
the Roman dominion in Asia.”
37. See the discussion of Avramea 1981 who determined the correct location of the
inscription.
38. For Amourios and Lamises, see pp. 109, 111.
39. Pachymeres X.16–21 (4.337–49), X.24 (4.355); Gregoras I.204ff.
40. Pachymeres X.23 (4.355–7).
41. Treiyer 2017 proposes that the date should actually be July 27, 1299, marshal-
ling a broad range of sources from the steppe to Cairo, some more convincing
than others. The subject deserves more detailed discussion than would be
appropriate here.
42. These will be discussed in Chapter 6.
43. Pachymeres X.25 (4.359–67).
44. Pachymeres X.20 (4.379f.), XII.1 (4.507). “Solymampax” may be Suleyman
paşa, emir of Kastamonu, but note the doubts of Failler 1994, 90f.
45. Pachymeres X.26 (4.369).
46. Patriarch Athanasius ep. 37, with the commentary of Talbot 345f. referring to
the parallel notice of Muntaner. This event was not recorded by Pachymeres.
47. Pachymeres X.29 (4.377).

48. As Zachariadou 1983, 6 supposes, but Pachymeres’ qualification of the Turks as
endotero suggests something closer at hand, perhaps Karesi whose settlement
and actions in this period are unknown.
49. Pachymeres XI.9 (4.425); for the identifications, see the magisterial discussion
of Failler (1994). The parallel text of Gregoras (I.214f) will be discussed
below, p. 164f
50. As argued by Failler 1994, 86f.
51. Beldiceanu-Steinherr 2000a, 429f.
52. See Zachariadou 1993b, 227, who, by emending the text of an inscription that
gives the family tree of the emirs of Karesi, proposes an identification with
Bagdi beg, ancestor of the dynasty.
53. Pachymeres XI.9 (4.425f.).
54. Pachymeres XI.16 (4.441f); cf. Foss 1976, 1299, 81–3.
55. For the confusing course of these events, see Korobeinikov 2014, 276–81; cf.
Beldiceanu 2003, 360f.
56. For this flood and its consequences, see Lindner 2006, 102–16.
57. The story of Ales Amourios is presented disjointedly by Pachymeres and is
not free from contradictions or loose ends; see Pachymeres X.20, X.25, XI.9,
XII.1 (4.349, 359, 363f., 425, 507) and the important discussion of Failler
1994, 96–104.
58. Pachymeres II.24 (1.185f.), III.3 (1.235), V.24 (2.515).
59. Pachymeres XI.16 (4.441f.) XII.1(4.503–7), XIII.13 (4.647).
60. For the Catalan campaign, see Pachymeres XI.14, 21, 23–6, XII.3 (4.437f, 463,
467–85, 527f.), and Muntaner 204–7. For its chronology, see Failler 1990, 84f.
61. For the disputed location of Germe, see the discussion of Jones 2014, 36–43.
62. For the location of Chliara, see Foss 1998, 160–6.
63. For the site and its remains see Foss 1979, 299–304.
64. Pachymeres XI.25 (4.475–9).
65. For the walls of Philadelphia, really about five miles long, see the comprehen-
sive discussion of Pralong (1984).
66. See below, pp. 172f., 181.
67. Pachymeres XI.24 (4.471).
68. Pachymeres XI.26 (4.483–5).
69. Pachymeres XII.3 (4.527), Muntaner 208–10.
71. Pachymeres XIII.4 (4.628).
72. Pachymeres XI.26 (4.481f.).
73. Pachymeres XI.27 (4.487f.). Pachymeres recounts another version by which
Choiroboskos escaped from captivity and went on to join imperial forces in
defeating Turks around Thessalonica.
74. Pachymeres XI.21 (4.453–7).
75. See Chapter 3, p. 00. This passage, if further proof were needed, demonstrates
that Belokome cannot be Bilecik, for the Ottoman heartland, long since
secured, was far from the scene of action here. For Katoikia, see Beldiceanu
2003, 370f.
76. Pachymeres XI.21 (4.455f).
77. Pachymeres XIII.9 (4.635f). For the site, see Hasluck 1906/07, 300, 306f., who
reported only “very scanty remains.”
78. Pachymeres XIII.17 (4.657); Monastery of Elegmoi: Janin 1975, 144–8; Mango
1968, 169–76.
79. Belke 2007. The three of these sites (except Rhodophyllon) that can be identi-
fied are all within ten kilometers of Apamea and thus not far from Elegmoi.
81. Pachymeres XIII.13 (4.647f.); for the date, usually considered to be 1304, see
Failler 1996.
82. Pachymeres XII.34 (4.609).
83. Pachymeres XIII.24 (4.681); cf. Kyriakides 2014.
84. Pachymeres XIII.25, XIII.35 (4.683, 701).
85. Pachymeres XIII.35 (4.701f.).
88. Miklosich and Muller 1860, I.80f.; Regestes du patriarchat V.62f. no.2086.
89. See the letter of Libadarios to Charles of Valois; Miklosich and Muller 3, 243f.
and the discussion of Ahrweiler 1965, 186–8 with references. Ahrweiler would
put the fall of the Hermus cities in late 1308.
90. For these dates, see Ahrweiler 1965, 43f., 47.
91. Ahrweiler 1983, 184, 190 with further references; Beldiceanu- Steinherr
1984, 17–22.
92. Gregoras I.361; Schreiner 1969, 389–95. For the local production, Beldiceanu-
Steinherr 1984, 29–34, Schreiner 1969, 411f.
93. Beldiceanu-Steinherr 1984, 22–9.
94. Gregoras I.384; date: ChronMin I.7.6, I.8.16.
95. ChronMin I.8.17.
96. Cantacuzene I.339f. By “Phrygia” Cantacuzene is anachronistically referring
to the region south of the Dardanelles and the Sea of Marmora, known in clas-
sical times as Hellespontine Phrygia, but more generally called Mysia.
Temirhan was actually the brother, not the son of Yahşı.
97. Cantacuzene I.341–60; Gregoras I.433–7; cf. Lindner (2007). For the forts of
the region, see Foss 1996a, 44–61.
98. Gregoras I.458; for the date: ChronMin. I.8.24, I.101.3.
99. See Foss 1996a, 24, 44.
100. Cantacuzene I.446–8, ChronMin I.8.27.
101. Cantacuzene I.459f. These two attacks pose a serious chronological problem,

for the Short Chronicle dates the payment to August 1333, so associated with
the event presented here first, for the attack of Cantacuzene I.459 involved no
negotiations, while that of I.446–8 specifically involved a place the emperor
had never visited before. It seems best, therefore, to presume a mistake in the
Short Chronicle.
102. For the fortresses of Izmir, see the comprehensive study of Muller-Wiener
1962, 60–96.
103. Umur is the hero of the Destan and the best-known Turkish leader of the age.
See the detailed study of Lemerle 1957, with 45–50 for Smyrna, and Inalcık 1985.
104. Lemerle 1957, 89–101. For the complex background of this alliance, see Laiou
1970. Umur’s raids and manpower will be discussed next, below, pp. 125–128.
105. For relations between Byzantium and the Turkish states in the 1330s and
1340s, see the sketch of Vismara 1968, which offers rather less than its title
suggests.
106. Cantacuzene I.482f.
107. Lemerle 1957, 102–15, with further references. The attack on Philadelphia is
narrated only in the Destan, but indirectly confirmed by Cantacuzene; cf.
Schreiner 1969, 396–401. An attack at this time was opportune, for the same
year saw the collapse of the Mongol Ilkhanid empire that had ruled Anatolia.
108. Cantacuzene I.480–2; Lemerle 1957, 110f., Zachariadou 1983, 38f.
109. Gregoras I.538–41; cf. the different account of Cantacuzene I.505f who

recounts that news arrived from Trigleia that a Turkish force from the eastern
cities that Orhan ruled, with thirty-six ships, was about to attack the neighbor-
hood of the capital that night or the next. Cantacuzene put together a force
that defeated the invaders after they had landed on the beach.
110. Gregoras I.545: For the date, usually given as 1337 see Gregortas tr. van Dieten,
II.2.385f., note 493.
111. Cantacuzene II.65–70, 77, 507.
112. Cantacuzene II.82.
113. Gregoras II.741–3; note the varying account of Cantacuzene II.552–6. There is
some confusion about the identity of this Suleyman: for Gregoras, he is satrap
of Troy but Cantacuzene refers to an unnamed “satrap of Lydia” in the same
context. Elsewhere (II.476), he mentions Suleyman “one of the satraps in Asia”
or (II.507): “son of Karasi satrap of Phrygia.” He appears in the Destan (2295)
as “son of Karesi.” The situation is complicated by the emirs of Aydın, Karesi,
and Saruhan all having sons named Suleyman: see Lemerle 1957, 204 n. 1.
Note that the coins of Suleyman of Karesi call him Suleyman b. Timur, i.e., son
of Timurtash: Ender 2000, 24. These must be given priority as an official pri-
mary source.
114. For Vatatzes, see Zachariadou 1993a, 231ff.
115. Cantacuzene II.589–92; cf. the rather different version of Gregoras II.762f.,
which Lemerle 1957, 220–3 considers inaccurate.
116. Karesi: this was Yahşı of Bergama, not his brother Temirhan of Balıkesir, who
maintained peace with Byzantium.
117. Lemerle 1984, 55–67, quoting a Greek text that gives considerable detail about
the fortifications. See the detailed study of Couroupou 1981.
118. Gregoras II.597f. This favorable view was not shared by the defenders of
Philadelphia in 1348, for whom Umur was a malignant figure, a worshipper of
the Devil: see the previous note.
119. Schreiner 1969, 401f.
120. Lemerle 1957, 236f.
3
Reconciling the Accounts
The Turkish and Byzantine sources—basically APZ and Pachymeres—

provide detailed accounts of the same periods, the same events, and the
same regions, though with very different approaches. Aşıkpaşazade focuses
on the Homeland, with an ever-broadening view as Osman and Orhan
extend their conquests into Bithynia and beyond. Pachymeres first notices
the Ottomans when they strike into Byzantine lands. But from 1302 onward,
Turk and Greek alike are presenting the same material. It is natural, then, to
hope to find cases where both recount the same event or place, so that the
sober Byzantine account can confirm or illuminate the colorful Turkish
one. There is no shortage of toponyms for comparison in both sources.
Unfortunately, though, such an effort is doomed to disappointment. Obviously,
the great cities—Bursa/Prusa, Iznik/Nicaea, and Izmit/Nicomedia—appear
in both narratives, in the right locations. But otherwise, there are precious
few examples of places or events common to both traditions, and where
they exist the Byzantine sources are often in direct contradiction with the
Turkish. A few long-favored identifications, based on apparent resemblance
of names, turn out to be incorrect:
Sagoudaous is not Söğüt. This place is mentioned only by Anna Comnena
(XV.ii.4) in association with Alexius Comnenus’ campaign of 1116 against the
Turks who were harassing Bithynia and Mysia. Some of the action took place
around the lake of Nicaea where Sagoudaous appears as near Fort St. George
and on the route that led down to the shore at Helenopolis. Even though its
exact location has not been determined, it had nothing to do with Söğüt.
Belokome is not Bilecik, because of phonetics and geography. Bilecik looks
as if it could be derived from Belokome, but the resemblance is misleading, for
the name of the Byzantine village would have been pronounced something like
Vilogume: B was pronounced V (as in Modern Greek) and so normally
appears in Turkish toponyms like Vize from Bizye. Initial B in Turkish
represents as Greek P as in Bursa from Prousa and many others. Also, -kome
(“village”) often turns into -güme, for which there are examples in western Asia
Minor. In other words, resemblance of these two names is illusory.
DOI: 10.1093/oso/9780198865438.003.0004
Pachymeres mentions Belokome twice, first in the context of other villages

evidently all in the same region: Angelokome, Anagourdys, Platanea, and
Melangeia, with Kroulla and Katoikia perhaps lying outside this cluster.1
Two of these places can be identified: Melangeia, better known as Malagina,
is an imposing fortress above the Sangarius, while Kroulla is modern Gürle
southwest of the lake of Nicaea. Katoikia has been—probably wrongly—
identified with Kite west of Bursa.2 In any case, these places are evidently in
maritime Bithynia, in the general region of the Sangarius and Nicaea, far
from inland Bilecik. In the second passage, Osman, returning from defeat-
ing a force that had been sent (apparently from Nicaea) to relieve Katoikia,
attacked and captured Belokome which had a garrison and a large treasure.3
Once again, the place is clearly in maritime Bithynia and could hardly be
Bilecik, long since conquered by Osman.
Angelokome is not Inegöl, whose name plainly derives from Aya Nikola,
presented as the tekfur of the city.4 The derivation is even more obvious
from the Ottoman name of the place, Aynegöl. Inegöl then was evidently
called Ayios Nikolaos, exactly like a fort of the same name near Philadelphia
which also became Aynegöl (though now called Sarıgöl).5 Naming forts
after saints (presumably the ones who gave them divine projection) seems
to have been common: Bithynia also had a Fort St. Gregory near the
entrance to the Gulf of Nicomedia and the Fort St. George near the lake of
Nicaea mentioned above.6 Pachymeres also mentions a Fort St. Elias near
Ganos in maritime Thrace.7 The phenomenon of giving the name of a place
to its ruler has parallels in APZ: Yalakonya whose name derives from the
toponym Yalakova, Kilemastorya from Kimasti, and Mihaliçi from the town
Mihaliç (now Karacabey).8 As for Angelokome, Anna Comnena (XIV.v.3)
mentions a river Angelokomites, but that appears to have been far to the
west, perhaps in the Troad.
On the other hand, there is one identification that is highly significant:
the Melangeia of Pachymeres is the Akhisar of APZ, the well-known for-
tress overlooking the Sangarius. Even more remarkably, both sources men-
tion the place in the same year, 1304, a rare confirmation of the seemingly
arbitrary dates that APZ occasionally provides—and perhaps of the extent
of Osman’s conquests in these critical years.
Normally APZ narrates exploits of Osman that are unknown to the other
sources, but there is one case where Osman seems to have made a spectacu-
lar move that does not appear in APZ. This is the raid that brought a Turkish
force to Scutari, directly opposite Constantinople in December 1302.9
As far as APZ was concerned, the first attack on the suburbs of Istanbul was
Reconciling the Accounts 137
the work of Akça Koca who took Samandıra and Aydos around the time of
the fall of Bursa.10
There is one very important event that appears in Pachymeres but not in
the Turkish tradition: the defeat of Osman in 1307 by a massive Mongol
force that drove him from the fortresses he had conquered from the
Byzantines and pushed his forces back into the recesses of Mount Olympus,
an event Pachymeres described as being more than one could hope for.11
This seems to have been decisive, for nothing more is heard of Osman—in
Greek or Turkish sources—for the next twenty years or so.
The battle of Bapheus in July 1302 has generally been treated as an event
of great consequence, so important that it must occur, under whatever
guise, in the Turkish narrative. Prof. Halil Inalcık made a brave attempt to
identify Bapheus with a battle mentioned in the Turkish tradition.12 He
used a text of Neşri dealing with Osman’s attack on Nicaea. According to
this, the besieged Nicaeans got a message out to the emperor, who sent a
force from Constantinople to relieve the city. They reached the promontory
of Dil, at the narrowest point of the Gulf of Izmit. Some had already crossed
the strait, about to follow the highway that led inland to Nicaea when they
were ambushed at night by the Turks. Those who had crossed were either
slaughtered or drowned; the rest retreated to the capital, leaving much loot
for the ghazis.13 The battle, such as it was, clearly took place in the vicinity
of Dil (north shore of the gulf) and Hersek (south shore). This in no way
suits the site of Bapheus, unambiguously described as near Nicomedia,
some sixty kilometers to the east.
This battle does not appear in APZ who, in cap. 16, gives only cursory
notice of the attack on Nicaea. On the other hand, the tradition represented
by the Anonymous (not by APZ), reports a similar situation around the
capture of Nicaea in 1331.14 In this version, the besieged people of Nicaea
managed to get a message to the “tekfur of Istanbul” who sent ships full of
soldiers to relieve the city. When they were disembarking by night on the
beach at Yalak Ovası, they were ambushed and slaughtered by the ghazis,
warned by a spy. In this case, the battle took place at Yalova, at the head of a
highway leading inland to Iznik or Bursa. As with everything to do with the
capture of Iznik and Izmit, the tradition is hopelessly confused.15
The importance, too, of Bapheus has been exaggerated. Note, for example,
the statement of Prof. Inalcık: “with the siege of Nicaea and his victory over
the emperor's relief army, Osman won incomparable fame and charisma
among the frontier Turcomans and leaders, securing for himself and his
offspring an enduring legitimation for primacy and sovereignty.”16 In fact,
the battle of Bapheus, where a large Turkish force defeated a small Byzantine
one, brought no great success for it was almost thirty years before Nicaea
finally fell to the Ottomans, and then to Orhan not Osman. Meanwhile, in
1307, Osman himself had been badly beaten by the Mongols and forced to
abandon most of his recent conquests and withdraw to the mountains. In
terms of his reputation, Pachymeres was certainly aware of him as a leading
enemy, but no more important than leaders of the other tribes or emirates.
As for Osman’s “fame and charisma,” his exploits certainly attracted a fol-
lowing from as far away as Paphlagonia and the Maeander but sources are
silent about his “enduring legitimation.”
Other major battles appear in only one source. Dimboz, where Osman
supposedly defeated a coalition of tekfurs in 1302, is absent from the
Byzantine sources. Likewise, the indisputably important battle of Pelekanon,
where a force led by the emperor himself withdrew in disarray, leaving the
remaining Byzantine lands in Bithynia nearest the capital open to Orhan,
finds no place in the Turkish sources, a most surprising omission.
Iznik and Izmit reveal a major problem with the tradition, for APZ
reverses the order of the conquest of these two great cities which repre-
sented an important stage on the rise of the Ottomans. The contemporary
Byzantine sources (with some support from the Arab traveler Ibn Battuta)
unambiguously state that Nicaea was conquered in 1331 and Nicomedia in
1338, dates no one disputes. In any case, Izmit, which could be supplied by
sea, would logically be the last holdout after Bursa, capable of being sur-
rounded on land, and Iznik, which could only rely on its lake. Furthermore,
the notion that Izmit was ruled by a woman related to the Byzantine
emperor suits the situation in Nicaea in 1306, when the emperor’s sister,
installed in the city, successfully defied Osman.
One detail, however, is common to both APZ who exalts in it and
Pachymeres who laments it: the conquest of the fort that was one of the
main defenses of Nicaea, called Karadin by the Turks and Trikokkia by the
Byzantines.17
All this suggests that APZ’s account derives from the distortions of an
oral tradition, which confused Iznik and Izmit, perhaps from similarity of
their names (Izmit was known then as Iznikmid) and has incorporated
memory of a situation that prevailed twenty years before the time of the
present narrative.
The momentous conquest of the coast of Karesi, which made the crossing
into Europe practicable, pose problems of a different kind, for it is impossible
Reconciling the Accounts 139
to reconcile the Byzantine and Turkish accounts. The Byzantine sources,

generally detailed for this period, mention the various rulers and divisions of
the emirate, but have nothing whatsoever to say of its actual acquisition—
whether by gift or conquest—by Orhan. APZ gives completely different
names for the rulers and seems ignorant of the divisions. His date for the
conquest—1335—is manifestly wrong, for texts and coins show that Karesi
was still functioning ten years later. The only safe conclusion to be drawn
seems to be that we know nothing of Orhan’s takeover of Karesi beyond the
fact that it happened. Once again, storytelling seems to have trumped history.
Comparing the Turkish and Byzantine sources produces very few exam-
ples of toponyms or events that can be seen from both points of view or
where one can illuminate the other. Plainly, the Homeland was too remote
from Constantinople to provide notice of a local chief who only attracted
attention when he burst onto the Byzantine scene after establishing a power
base in interior Bithynia. There is therefore no way to judge the accuracy or
veracity of APZ’s account of the rise of Osman before he appeared in the
vicinity of major Byzantine cities, but when he did—and information is
available in both sets of sources to invite comparison—the results are not
encouraging. APZ has confused Iznik and Izmit, as he has confused the two
Ottoman attacks on Nicaea: failed in 1307, successful in 1331. Here, as in
the case of Karesi, an oral tradition seems to have woven stories around
events a century and half before APZ’s own time. Hope of understanding or
even establishing the course of Osman’s rise in the Homeland seems unlikely
to be fulfilled.
Historians of the Seljuks of Rum, writing in the early fourteenth century
and well acquainted with Asia Minor, might be hoped to provide informa-
tion or perspective on Osman and his lands, but for them too he was too
obscure or remote to attract notice. None of the three major sources—ibn
Bibi, the Anonymous History of the Seljuks of Asia Minor, and Aksarayi—
ever mentions Osman or, for that matter, Aydın, Saruhan, or Karesi, though
Karaman and Eşref frequently appear along with the states based on Afyon
Karahisar or Denizli, with Kastamonu, Menteşe, and Germiyan making
occasional appearances.
In other words, the Byzantine and Turkish sources rarely present the
same information, which could be validated by comparing the differing
accounts, but they are rarely in direct contradiction. Ideally, they could be
used to supplement each other, making allowance for the oral—and there-
fore potentially inaccurate—nature of the Turkish tradition.
Notes

2. Beldiceanu 2003, 371.
4. Already noted in passing and in a different context by Babinger 1922, 151.
Angelokome appears only in Pachymeres XI.21 (4.453) the identification was
made by Texier in the early nineteenth century and universally followed: a
signpost on the road entering İnegöl proclaims the name Angelokome. See
Ramsay 1890, 206f, who considers the problems the identification raises and
offers a most implausible solution.
5. Schreiner 1969, 390 n. 1.
6. St. Gregory: Failler 1990, 21f.; St. George: Giros 2003, 215; for its approximate
location, Lefort 2003, 465.
8. APZ cap.30, 35.
10. APZ cap. 25–6.
12. Inalcık 1993 and, in more detail, 2003.
13. Translated by Inalcık 1993, 85f.
14. Ibid., 83f., a passage translated by Prof. Inalcık from a later source.
15. Prof. Inalcık does not help his case by treating the two attacks on Nicaea
(1306–1307 and 1331) as if they were one, often rendering his arguments hard
to follow.
16. Inalcık 1993, 97.
17. APZ cap. 22, Pachymeres XIII.35 (4.701f.).
4
Non-Narrative Sources
Coins
The written sources combine to populate the realistic landscape that

Aşıkpaşazade (APZ) presents, but they are not the only material that might
illuminate the still obscure rise of the Ottomans. Coins, which constitute
contemporary hard evidence, could potentially give some insight into the
wealth (or poverty) of the times as well as patterns of trade and relations to
superior powers. In this case, their contribution is real but limited, for
important aspects of the numismatic record are missing: hoards, which
could illustrate patterns of trade and fluctuating degrees of prosperity; and
excavated coins, which could illuminate local economic life. Nevertheless,
coins can provide some insights when the Ottoman issues are seen in the
context of those of the western emirates.1
For Islamic states, striking coins is a significant act, because coinage
(sikke) along with mention of the sovereign’s name in the sermon in the
Friday mosque (hutbe) was a sign of independent sovereignty. According to
APZ, Osman, after settling his new conquest, Karaca Hisar, was faced with a
popular request for a Friday mosque and a religious judge (kadi). Their can-
didate, Dursun Fakih, however, pointed out that this required the permission
of the Seljuk sultan. Osman replied that he had conquered the land by him-
self and had no need for any sultan. Dursun thereupon became kadi and led
the Friday prayers in which the hutbe was read. This supposedly happened in
699/1299 and has been taken to mark the beginning of Ottoman history.
According to APZ (cap. 8), the sultan had already rewarded Osman for
his capture of Karaca Hisar with a banner, tent, weapons, and animals in
687/1288–1289. Neşri combines the two events: although the sultan’s gifts
effectively recognized independence, Osman chose instead to recite the
hutbe and strike coins in the name of the sultan.2 These coins, if they ever
existed, have not survived. Later, when the sultan died, Osman named
Dursun Fakih as kadi and prayer leader.
DOI: 10.1093/oso/9780198865438.003.0005
These narratives may have a kernel of truth, for Osman had the laqab or
honorary name Fakhr al-din, attested in a document of 1324.3 Since these
titles were normally conferred by a higher authority, such as the Seljuk sultan,
its existence may indicate Osman’s receiving some reward or recognition from
the sultan. In any case, the sultan was not the famous Ala ad-din Kayqubad
I, who died before Osman was born, but if any of that name were involved, it
would necessarily be Ala ed-Din III, who reigned intermittently (1284,
1293–1294, 1301–1303). He was an insignificant puppet of the Mongols,
part of a facade that involved coins being struck in his name. In a comparable
case, the equally powerless Giyath ed-Din Kaikhosraw III (1265–1282) figures
in the will of ibn Jaja with a multitude of extravagant titles.
The sources imply, but do not state, that Osman stuck coins in his own
name. In fact, no coins that were certainly issued by Osman have been dis-
covered. There are, however, two pieces that ostensibly were products of
Osman’s rule. The first, known since the 1970s, weighs far too little for a
normal akçe (the small silver coin of 1.15g of 900 silver that was the stand
ard Ottoman issue); the second, only recently published, has been identified
as a half-akçe, though it seems heavy for such a denomination (which was
hardly ever produced in this period).4 The first coin, which bears the name
“Osman ibn Ertughrul” on both sides but no place of mintage (parts of the
inscription cannot be read), is controversial. Of the second, struck from the
same or similar obverse die, the reverse inscription cannot be read at all.
The three experts of this coinage have divergent opinions: Atom Damali
considers the first genuine, but according to Slobodan Srećković it differs
in weight, design, calligraphy, and inscription from the standards of the
period. Rolf Ehlert, on the other hand, assigns both coins to Germiyan on
the basis of an ornament that appears on the first coin and on a recently
discovered anonymous piece of that principality, noting that some akçes of
Germiyan are seriously underweight.5 These anomalies suggest that these
coins are not regular products of Osman, but neither their date nor the
occasion nor reason for their production have been proposed or determined.
Another coin has been attributed to Osman by implication.6 This is an
akçe bearing the name of the Ilkhan Ghazan, the date 699, and a mint that
has been read as “Söğüt”—i.e., ostensibly struck in the town where the
Ottomans had their origin and at a time when they had moved their capital
to Yenişehir near the Byzantine frontier. The obverse legend—the kalima,
the Muslim profession of faith—is inscribed within a square, while obverse
and reverse alike are limited by a beaded circle. It differs in design from the
main empire-wide coinage of Ghazan but resembles some regional issues
Non-Narrative Sources 143
struck in southern and southwestern Anatolia.7 The mintmark consists of

the letters sin-àyn- dal: ‫ سعد‬which could be read as S.Gh.D, i.e., Söğüt. The
problem here is that that is not at all the way Söğüt is spelled in Ottoman
Turkish, where it appears as ‫ سكود‬SGUD, the spelling found in the manu-
scripts of APZ and in endowment documents going back to the fifteenth
century.8 There is no way of mistaking this word for the one that appears on
the coin, which must therefore be regarded as an imitation with a garbled
inscription, rather than any evidence for Osman.
In other words, Osman, unlike some of his neighbors, did not strike
coins in his own name or in the name of a Seljuk or Mongol sovereign, for
no issue of theirs was apparently ever produced in lands under Osman’s
control. This has implications for the economy of the nascent Ottoman state
and for its wealth—and perhaps for Osman’s claims to independence. It
suggests that the Ottoman enterprise was relatively poor or at least not very
developed economically.
The reign of Orhan brings major change. He not only issued coins but in
a variety of types and inscriptions not seen again until the reign of Mehmet
Fatih a century later. There are eight (or eleven or six, depending on the
catalog) main types of the akçe with several varieties.9 The majority bear the
mint name Bursa; no other mint appears. All agree that the first type is the
akçe in an Ilkhanid style bearing the mintmark Bursa and the date 727
(1327)10 (Fig. 4.1). It shows that Orhan was asserting his independence and
Fig. 4.1 Orhan’s new coinage, struck in Bursa in 1327. © Classic Numismatic
Group, LLC, http://www.cngcoins.com, reproduced with permission
producing a sophisticated well-designed coinage already only a year after

capture of his first major city, Bursa, which became his capital.
One undated series gives Orhan the grandiose title al-sultan al-a`zam
“the greatest sultan” while another calls him al-sultan al-adil (the just sultan),
and a third gives both titles.11 One variant of the al-sultan al-a`zam type is
the most common of Orhan’s issues; it has been dated to the 740s/750s.12
These titles are in striking contrast with the coins of his two successors, who
built a mighty empire half based in Europe, and were far more powerful
than Orhan ever was. Both Murat I and Bayezid I modestly appear on their
coins with name and patronymic alone.
Orhan’s titles find parallels in the emirates: Yahşı Han of Karesi also
called himself al-sultan al-a`zam; Yakub I of Germiyan appears to have
used the same title. On one issue Yahşı appropriates amir al-a`zam while on
another he calls himself merely bey. His brother Demirhan claims to be
malik; Mehmet of Aydın is more grandly al-sultan al-malik al-jalali. Bayezid
of the Candaroğulları is most often called al-sultan al-a`dal.13 In other
words, Orhan’s claim to the title of sultan is not all unusual in this context.14
Ottoman coinage, which begins in 1327, is roughly contemporary with the
coinages of the western emirates of Karesi, Aydın, Menteşe, and Germiyan
where a date in the 1320s or 1330s seems normal for the beginning of
independent coinage. This, of course, coincides with the weakening of the
Ilkhanids, and their collapse after the death of Abu Saìd in 1335. In most
cases, there is a long gap of twenty years or more between the establishment
of the emirate and the first independent coinage. To some extent, issues in
the name of the Ilkhan ruler seem to fill the gap: Germiyan, Karesi, Sasa Bey,
Aydın, and Candaroğlu struck such coins, to which Menteşe, issuing in the
name of the Seljuk sultan, may be added.15 These were mostly produced
around 1300 and in small quantities, but they raise important questions
which can be approached by considering the coinage of the ephemeral but
well-attested Sasa Bey.16
Sasa, who is mentioned by two Byzantine historians as well as the Turkish
epic, the Destan of Umur of Aydın, and the contemporary Catalan chronicle
of Ramon Montaner, appears in history in the spring of 1304 when the
Catalan army in mercenary service of Byzantium drove his forces and those
of Aydın away from attacking Philadelphia. By the end of the year, however,
the Catalans were recalled to Europe and the Turks returned. On October
24, 1305, Sasa took Ephesus.17 Pachymeres identifies him as the son-in-law
and therapon (attendant) of Karmanou Mantakhiou, i.e., Menteşe of
Germiyan. According to the Destan, Sasa took the mountain fortress of
Birgi, then called in the sons of Aydın, an action described as opening the
gates of conquest for Germiyan, whose stirrup they held. Then, together,
Sasa and Aydın captured Ayasuluk/Ephesus, where they converted churches
into mosques, then conquered Keles/Kaloe. After the sons of Aydın defeated
a “Frank” counterattack, Sasa defected, joined the Christians, and was killed
in battle. For the Destan, the scene of action is entirely in Ephesus and its
hinterland, the Cayster valley.
Gregoras mentions the “Persarchos” Sasan as ruling a coastal district
commanding the mouths of the Maeander and Cayster rivers. In any case,
the forces of Aydın soon took over, for its founder Mehmet, in the building
inscription on his Friday mosque in Birgi, proclaims that he had conquered
that city in 707 (1307/08), suggesting that the career of Sasan lasted five
years at most.18 He had been closely associated with Aydın and Menteşe and
was responsible for the conquest of the fertile Cayster valley and the impor
tant commercial center, Ephesus.
Numismatics may cast further light on these events. Sasa issued no coins
in his own name but the mint of Ayasuluk (Ephesus) struck in the name of
the Ilkhan sultan Uljaytu in 706 (1306/1307).19 Similar pieces were issued in
these years in the lands that were to form the beylik of Aydın: Ayasuluk in
710; Sultanhisar 705 and 710; and Tire in 707.20 None of these name the
local emir, only the Mongol sovereign.
At first sight, a Mongol ruler seems completely out of place here, for
although the Ilkhanids were in full control of the Anatolian plateau, there is
no evidence that they ever set foot in this region: the sources make it very
clear that its conquest was the work of Sasa and the sons of Aydın. The two
contemporary authors, Muntaner and Pachymeres, could hardly have failed
to mention anything so extraordinary as the appearance of a Mongol force
on the Aegean coast. In fact, the Mongols very rarely intervened directly
into their Wild West—the land of the Turcoman tribes—and then only to
suppress revolts.21
In other words, there is no reason to take the Ayasuluk coin as represent-
ing a Mongol conquest or Mongol direct rule over this maritime district. It
was produced within a year of Sasa’s conquest of the city and no doubt
under his authority. This raises questions about the nature of Sasa’s power
and that of the Ilkhanids. Normally, if a coin bore the name of the Ilkhan
and were struck, say, in Tabriz or any of a range of mints in Iran or Anatolia,
it would be taken to indicate direct rule, with the branch mints striking the
same types, with the same metrology as the capital. Since that cannot be the
case here, this coin must represent a different relation between the Mongols
and the local ruler, who is evidently recognizing the overlordship (in this
case of Uljaytu) whether because he was acting directly in the Ilkhan’s inter-
est or proclaiming a nominal sovereignty or some other relationship.
The Egyptian statesman and historian Al-Umari, writing in the 1330s,
provides a clue. He explained that when the Mongols took over, the Seljuk
sultans were left with titles but no authority. Real power was in the hands
of the Mongol governor; the Friday prayer was made in the name of the
princes descended from Genghis Khan; gold and silver coins were struck
in their name. Later, when the Seljuks had disappeared and the Turcoman
states had established their independence, they didn’t stop seeking the
goodwill of the Mongols. They sent rich gifts, maintained agents at the
Mongol court, said prayers in the name of the Mongol ruler of the house
of Hulagu (i.e., the Ilkhans), and struck coins in his name.22 All this in an
effort to gain Mongol support (or at least acquiescence) for their inde
pendent existence and possibly advantages over rival emirs. In this con-
text, it appears that Sasa was placing himself under Mongol protection or
was actively seeking their goodwill. In any case, the Mongol name was
making its appearance in a land far from the territories they ruled—and
not only in Ionia, other emirates also name Mongol rulers on coins struck
in their capitals:
Kutahya, capital of Germiyan, struck in the name of Ghazan in 698
and 700.23
Bergama, capital of Karesi, in the name of Uljaytu (703–16)24

Kastamonu, capital of the Candaroğlu, in the name of Ghazan (701), Uljaytu
(716), and Abu Sa’id (723–8)
Note that three of these (Germiyan, Karesi, and Kastamonu) are Osman’s
neighbors and that the name of the Mongols, exerted influence if not direct
power, close to the lands of Osman who differed from the others in not issu-
ing coins at all.25
Inscriptions and buildings
Orhan’s titles sultan al-a`zam and sultan al-adil may seem at first sight
unexpected and extravagant, but as noted, they are part of the common lan-
guage of sovereignty in the fourteenth-century beyliks. Actually, the small
size of the coins allows for few honorifics; the titles on them are nothing
compared with what appears in monumental building inscriptions of the
age. One of the grandest examples adorned a mosque in Bursa.26 Dated 738
(1337), the titles it contains have given rise to much discussion:27
al-amir al-kabir al-mùazzam al-mujahid fi sabil Allah sultan al-ghuzat

ghazi ibn al-ghazi shuja` al-dawla wa’ldin wa-l afaq bahlavan al-zaman
Urkhan binÙthman
“the Exalted Great Emir, Warrior on behalf of God, Sultan of the Ghazis,
Champion of the State and the Religion and of the Horizons, Hero of the
Age, Orhan son of Osman”
This is one of the only two inscriptions that name Orhan; the other, dated
740/1339, on his mosque in the bazaar of Bursa, is more restrained:28
sultan al-ghuzat wa’l-mujahidin Orhan beg bin Osman beg

“sultan of the ghazis and warriors for the faith, Orhan bey, son of Osman bey”
Contemporary examples from other beyliks will provide a context for

Orhan’s titles and show how bombast and restraint coexisted.29
Germiyan, the most powerful of the emirates, seems to show the greatest
moderation:
al-
amir al-ajall al-
kabir Yakub ibn Alishir (Ankara 699/1300: RCEA
13.5080)
“the great, most exalted amir, Yaqub son of Alishir”
al-amir al-ajall al-kabir sultan al-Kermiyaniya (Sandikli 725/1325: RCEA
14, 5517)
“the great, most exalted emir, sultan of Germiyan”
though one of their chiefs makes a seemingly unusual claim:
malik al-umera wa’l-kubera mubariz al-din Umur b Savji.
“King of the emirs and the great, Fighter for the faith” (Kutahya 714/1314:
RCEA 14, 5346)30
Urkhan of Menteşe stresses the jihad:
al-amir al-muàzzam al-mufakkham al-mansur al-muzaffar sultan ghuzat

al-atrakshuja al-dawla wa’l-din Urkhan ibn Masud.
“Exalted, honored, victorious, triumphant amir, sultan of the ghazis of the
Turks,Hero of the State and the Religion, Orkhan son of Masud” (Peçin
732/1332: RCEA 15, 5622)
His son is more restrained:
al-
amir al-kabir al-ajall Ibrahim beg ibn Urkhan (Muğla 1344: RCEA
15, 5983)
“the great, most exalted amir, Ibrahim bey son of Urkhan”
Mehmet Aydınoglu’s building inscriptions are relatively moderate:
mawlana al-amir al-kabir al-ghazi fi-’l sabil Allah Muhammad ibn Aydın
“Our lord the great emir, ghazi on the path of God” (Birgi 712/1312:
RCEA 5310)
al-jenab al-àli al-amir al-kabir al-alim al-adil abu’l-khair mubariz al-dawla
w’al-din Muhammad b Aydın.
“Exalted, honorable, the great, learned, just amir, founder of pious works,
Fighter forthe nation and the religion” (Birgi 712/1312: RCEA 5311)
His epitaph is bombastic:
al-amir al-kabir al-alim al-mujahid al-murabit abu’l-khair sultan al-ghuzat

mubarizal-dawla wa’l-din Muhammad b Aydın.
“The great, learned amir, fighter and defender (of the faith), founder of
pious works,sultan of the ghazis, Fighter for the nation and the religion”
(Birgi 734/1334: RCEA 5657)
Inscriptions such as these, which celebrate construction or repair of

mosques (the great majority) or other works of public utility, can be com-
bined with others from the region and the time to suggest a rough guide to
the prosperity of the beyliks. Construction of a mosque was a primary need
once a city had been conquered and settled. These were often small scale, to
be replaced later by substantial Friday mosques, where the congregation
gathered and the hutbe was proclaimed. They were built only in towns and
cities and enjoyed a special status. They were often on a grand scale, adver-
tising the glory of the religion and the wealth of the ruler.31
Known examples reflect a wealth that was increasing in the fourteenth
century. The earliest was the Ulu Cami or great mosque of Birgi, built in
1312 soon after the conquest (Fig. 4.2). Others called Ulu Cami or Friday
Mosque are known from inscriptions or mentioned by Ibn Battuta:
Kastamonu (Candaroğlu) before 1332

Balikesir (Karesi) completed after 1327
Peçin (Menteşe) 1332
Fig. 4.2 Birgi’s great mosque (1312)
Sinop (Candaroğlu) 1341

Muğla (Menteşe) 1344
Menemen32 (Saruhan) 1358
Ayasuluk (Aydın) 1375
Manisa (Saruhan) 1376
Sandikli (Germiyan) 1378
Bursa (Ottoman) 1396
Bergama (Ottoman) 1398
In other words, substantial construction was underway throughout the

fourteenth century, with ever-grander buildings. The most splendid in size,
decoration, and innovative architecture were the Isa Bey Cami of Ayasuluk,
the Ulu Cami of Ishak Bey of Saruhan, and especially the Ulu Cami of
Bursa, built by Beyazit I long after the capital had moved to Europe.33
Prosperity was evidently increasing during the first century of Turkish rule,
and the Ottomans were participating, though perhaps at a slower pace.
A more impressive indicator of prosperity is the foundation of entirely
new towns which functioned as capitals of emirates. Best known is Peçin in
Menteşe, just five kilometers from the former capital, Milas. When Ibn
Battuta visited in 1331, the city, together with its congregational mosque
was under construction. It soon came to have a full complement of religious
and secular buildings.34 A few years earlier Karesi bey, who died in 1327,
had built his new capital, Balıkesir, described by Ibn Battuta as a “fine and
prosperous city,” but the new congregational mosque was still lacking a roof.
Osman was also one of this company, for he founded his new capital,
Yenişehir between 1299 and 1302 according to Aşıkpaşazade’s dubious
chronology. Proclaiming the wealth and success of ruler and religion may
have been a prime motive of Menteşe and Karesi but the site of Osman’s
Yenişehir suggests an additional reason, for the city is over a range of high
hills from Nicaea, some twenty kilometers away, and in easy striking dis-
tance of Bursa and the Sangarius valley, a good location for an emir poised
to move against the declining empire of Byzantine Anatolia.
Yenişehir preserves the remains of a few buildings attributed to Orhan,
among them a mosque, a medrese, a dervish lodge, and a hamam, but noth-
ing that can certainly be attributed to Osman.35 The picture is similar
throughout the entire Homeland, where there are thirty-six mosques built
by Orhan or attributed to him. Several are in small villages, most have been
completely rebuilt.36 The most substantial are in cities that figure in
Aşıkpaşazade: Bilecik, Bursa, Izmit, Iznik, Söğüt, Yar Hisar, Yenişehir. Iznik
contains several other buildings datable to Orhan’s reign, while Bilecik has
an imaret and a turbe. Bursa saw an outburst of construction: mosques,
schools, dervish lodges, soup kitchens, caravansarays, tombs, baths, fortifi-
cations, palaces, bridges, and many others.37 These completely transformed
the city. Söğüt contains a tiny mescit supposedly built by Ertuğrul.38 It has
been extensively rebuilt, but the town has nothing of Osman who is con-
spicuous by his absence through the whole area. A bath in Yenişehir may
date to his time, but the Osman Cami in Bilecik was built by Orhan in
honor of his father, and the existence of his mosque in Karacaşehir, attested
by an endowment document, seems dubious.39 The buildings, like the coins,
reflect the great difference between the realms of Osman and Orhan.
Documents
In a groundbreaking study of archival vakf (endowment) documents from

the Homeland area Mme Beldiceanu was able to gain some important
insight into the comparative resources of Osman and Orhan.40 In the mid-
fifteenth century register of the district of Sultanoyugu, Osman is men-
tioned only three times (Orhan appears twice as often). Where Osman
appears, the endowments are of small farms, apparently granted in exchange
for military service. The registers of Hudavendigar, the province of Bursa

(1456 and 1521) mention Osman only in Söğüt, Yar Hisar, and Ermeni
Pazar; Orhan appears in every part of the province. Again, these are dona-
tions on a small scale, nothing capable of supporting a mosque or school or
hospice. The situation changes drastically when the focus moves north to
Bursa and Iznik, and the fertile plains and valleys of the Mediterranean cli-
mate area. Naturally, Osman does not appear at all, but there are numerous
references to Orhan whose endowments are on a large scale capable of sup-
porting public buildings.
Osman is named in endowments of the fifteenth and sixteenth centuries
that reveal activity on a smaller scale. They are usually of simple farms, that
show him in possession of parts of the Homeland, with five from Söğüt, two
from Yarhisar, and one from Ermeni Pazar. Most of the villages concerned
cannot be identified, but one endowment of Osman was situated at Matlap
(modern Muttalip) less than ten kilometers north of Eskişehir.41 This is
important information, being the first independent confirmation of
Osman’s presence in a specific part of the Homeland. This place has further
significance and raises a central question, viz., Osman’s relation to the
Mongols, for Muttalip was on or adjacent to the land endowed by Jebrail ibn
Jaja in 1272. At that time, only a decade or so before Osman supposedly
appeared on the scene, Eskişehir and its surrounding villages were under
Mongol control; now Osman appears to be established in the district. The
implications of this will be considered in Chapter 6.
So far, the evidence suggests that the Homeland was relatively poor and
that Osman (and even Orhan before the conquest of Bursa) could derive
little profit from the lands he controlled and was in no position to support
monumental building. However, a detailed endowment document issued by
Orhan in March 1324—the oldest surviving complete and genuine Ottoman
document of its kind—presents a rather different picture.42 In it Orhan
grants the entire district of Mekece for the foundation and maintenance of a
lodge (zaviye) for dervishes, travelers, and the poor. This is evidently a sub-
stantial foundation that has major implications.
Mekece stands adjacent to the Sangarius at a point where two major
roads meet: the west–east route from Iznik to Kastamonu and Sinope and
the north–south route along the Sangarius. The former was followed in 1332
by Ibn Battuta, who spent the night in Mekece, but in the house of a legist.43
This meant not only that Orhan in 1324 disposed of valuable property, but
that the roads were secure enough that travelers of all kinds could pass.
At this time, however, Iznik was still in Byzantine hands. This document
implies that Orhan controlled the route east from the city, presumably as
part of the systematic blockade that Osman had attempted in 1304–1307.
As for the Sangarius route, its most important bastion for Byzantium had
been the powerful fortress of Malagina. Here Pachymeres and APZ offer a
rare confirmation of each other: the Byzantine reports that in 1304
“Melangeia” was emptied of its inhabitants, while the Ottoman recounts the
capture of Akhisar in 704.44 Osman probably held this conquest for a very
short time, for the defeat the Mongols inflicted on him in 1307 forced him
to abandon the forts he had conquered and to withdraw to Mt. Olympus
above Bursa. Yet the Sangarius route was evidently in the hands of Orhan in
1324, as it presumably had been for some time, long enough for the region
to be pacified, for traffic to resume and for resources to accumulate.
In other words, Osman had been responsible for these gains, for 1324
appears to mark the beginning of Orhan’s reign, though strictly speaking
the date of his accession and of the death or incapacitation of Osman is
unknown.45 This document, therefore, can be seen as reflecting conditions
in Osman’s time and suggesting that he was richer and more successful than
other evidence would seem to indicate. The document is written in Persian,
the language of the Seljuk chancery, and employs their elaborate script. This
is evidence for an unexpected degree of sophistication and an organized
state. Even more unexpected because it was issued before the conquest of
Bursa and therefore in the obscure Yenişehir. The Ottomans were no longer
simple nomads, but at least in this case, had formed something resembling a
traditional Islamic state.46
The Mekece document also raises questions about the structure of
Ottoman rule. That the names of four of Orhan’s brothers appear as wit-
nesses to the endowment has suggested that Orhan presided over a family
enterprise where the chief was elected and conferred with members of the
family for making important decisions.47 Actually, Orhan, like Osman
before him and his descendants and successors, ruled by himself with no
power-sharing or challenge to his authority. APZ (cap. 29) relates that after
Osman died, Orhan’s brother Ala ed-Din and other leaders met to deal with
Osman’s modest possessions. On being asked his opinion, Ala ed-din stated
that the flock only needed one shepherd and that should be Orhan. He
requested only for himself a farm near Kite, which was granted. Although
stories about him are largely mythical, Ala ed-din was a real and evidently
important figure.48 Orhan had at least four other brothers, but there is no
evidence that Ala ed-din or any of the others played a significant role in
running the state. The only occasion where one of them, Pazarlu, appears in
history, at the battle of Pelekanon, it is in a subordinate role.49
Notes
1. These are discussed in more detail in Foss 2019.
2. Neşri I.108.
3. Beldiceanu-Steinherr 1967, 86 n. 8. See the article “lakab” by C. E. Bosworth.
4. First coin: Srećković 1999, 11–13; Damali 2010, 95, Ehlert 2014, 14f; second:
Ehlert 2014, 13.
5. Srećković 1999, 11f; Damali 2010, 95; Ehlert 2014, 15 (see his entire discus-
sion 5–18).
6. Lindner 2007, 97f., illustrated.
7. Standard types: Diler Ga281–Ga290; regional: Diler Ga291–Ga294.
8. See Hudavendigar Livası Tahrir Defterleri I. 267–92.
9. Remler 1980, 181–5; Damali 107–29; Srećković 15–28; Ehlert 5–50. It is
extremely difficult to reconcile these types since each author seems to include
coins missing from one of the others.
10. Damali G6a p. 117; Ehlert type 1a, p. 30f. The coin bears a general resemblance
to the contemporary issues of Abu Sa’id (1316–1335) but differs in detail from
all of them.
11. “Greatest sultan”: Damali G7–G10 pp. 118–25; Ehlert types 3–5 pp. 36–8; “just
sultan”: Damali G12 p. 126; both titles: Damali G11 p. 125.
12. Ehlert type 7, p. 43f.
13. Many more examples could be found among the coinage of another set of
Turcomans, the Artuqids, and Zengids of northern Syria and Iraq, who flour-
ished in the previous century. See Elisseeff 1954.
14. Lowry 2003, 37, 43, evidently unaware of these coins, claims that Murat I in
1388 was the first to call himself “sultan.” For the meaning of the term “the
power that one exercises and by extension the chief who exercises it,” see
Beldiceanu 2002, 227f. with further references.
15. These issues will be discussed in the next chapter. Menteşe, Aydın, and Saruhan
also produced western-style silver coins in the 1330s.
16. Note that the Byzantine sources call him “Sasan,” while the Turkish use the
form “Sasa.”
17. Pachymeres XIII.13 (4.647f.); for the date, usually considered to be 1304, see
Failler (1996).
18. RCEA 5310. The inscription raises chronological problems that resist solution:
it has Mehmet conquering Birgi in 707 while the Destan attributes this to Sasa
before the Aydınoğulları arrived and likewise (line 7) states that Mehmet only
became emir in 717. See the discussion of Lemerle 1957, 19–25.
19. Izmirlier 2005, 107: it does not name the sultan, but only gives his titles in
abbreviated form.
20. Ayasuluk: Izmirlier 2005, 108; Sultanhisar: ibid. 108–10; Tire: ibid 117–19.
Yılmaz Izmirlier (2005, 45–51) identified this group of coins as associated with
Sasa bey, noting that they share the same metrology and design: see his
catalogue ibid. 107–10, 117–19. The same author also presents coins suppos-
edly from Arpaz and Karaağaç but these require confirmation, since those of
“Arpaz” seem to name the Seljuk Alaeddin—the reading is uncertain—while
Izmirlier reads the “Karaağaç” issues as “Karağaç rekz,” which he interprets as
“hoisted the flag.” Equally questionable are the issues of Tire bearing the term
transliterated as “ghuzzad”: ibid., 119–21.
21. They intervened in the coasal regions in 1262, 1292, 1307, and 1327:
Korobeinikov 2014, 223f., 23; Pachmeres XIII.37 (4.709); Beldiceanu 184, 22–9.
22. al-Umari 49–52/374–7.
23. 698: Kütahya (Diler 2006, 359);700: Kütahya (Ender 2005, 25).
24. Diler 2006, 410; the coin does not bear a date.
25. In the 1330s, three of the maritime emirates—Menteşe, Aydın, and Saruhan—
issued western-style silver coins which will be discussed below, p. 195.
26. Mantran 1954, 89 no. 1. It is out of place on the Şehadet Cami where it now
appears, for that was only built in 1389. It may have belonged on the destroyed
mosque of Orhan in the citadel: Ayverdi 1966, 58f.
27. It forms the basis for the “ghazi thesis” advocated by Wittek 1936; see
Lowry 2003, 33–44. The readings wa-l afaq, al-zaman are uncertain; see the
text, transcription, and translation in Lowry 2003, 34–8. For the meaning of
Orhan’s titles, especially sultan, ghazi, and mujahid, see Beldiceanu 2002.
28. Mantran 1954, 90 no.2. These are not the earliest Ottoman inscriptions; the old-
est is on the Haci Özbek mosque in Iznik, dated to 1333: Otto-Dorn 1941, 15–18.
29. Emecen 1995 pointed out the significance of the inscriptions of the beyliks.
30. Actually, malik had lost much of its original meaning and come to denote
someone as “head of ” or “subordinate ruler of ”: see the article “malik” by
A. Ayalon in E!2 and Lewis 1988, 53–6.
31. See Necipoğlu 2005, 55–7.
32. Beldiceanu 2015, 283.
33. Note that this great conqueror merely calls himself al-sultan al-muàzzam Bayazid
khan ibn Murad khan, much as on his coins: Mantran 1954, 91 no.3 dated 802.
34. Description, plans, and illustrations in Arel 1968.
35. Ayverdi 1966, 206–16.
36. Listed in Ayverdi 1966, 8–20 and described in the following chapter.
37. Ayverdi 1966, 49–118.
38. Ayverdi 1966, 2f.
39. Ayverdi 1966, 15, 34f., 13. Karacaşehir poses special problems: see below,
pp. 168–170.
40. Beldiceanu 2000b.
41. Ibid., 24. Other villages that can be identified (Günyarık, Aşağı Söğüt, Sevinç,
Çukurhisar) are on the road from Söğüt to Eskişehir and virtually surround
that city, but it is not clear whether they are sites of foundations of Osman or of
one of his successors.
42. Beldiceanu 1967, 85–9; Lowry 2003, 72–9 with text, transcription, and
translation.
44. Pachymeres XI.21 (4.455); APZ cap. 20.
45. Beldiceanu 2015, 235f. = Lowry 2003, 74. See the Appendix in Chapter 6.
46. This point is made by Lowry (above, n. 42).
47. Beldiceanu 1967, 87, 97f. The notion of a collective enterprise may reflect later
nostalgia for a time when sultans didn’t kill off their brothers: cf. ibid. 87 n. 16.
48. For Imber 1993, 68–71 he is entirely fictitious, but his 1333 endowment shows
that not only did he exist but he possessed considerable property: Beldiceanu
1967, 94–9.
49. Cantacuzene I.349.
5
The Overlords
In the time of Osman, Asia Minor was at the center of a ring of contending
powers: the Mongol Ilkhans of Iran in the east, the Mamlukes of Egypt in
the south, the Mongol Golden Horde north of the Black Sea, and the ever-
diminishing Byzantine state in Thrace, Macedonia, Greece, and western
Anatolia.1 The Ilkhans and Mamlukes were bitter enemies, as were the
Ilkhans and the Golden Horde. Byzantium, which occupied a strategic location
on the trade routes between eastern Europe, western Asia, and North Africa,
maintained a delicate diplomatic balance amid the great power rivalries. By
this time, the former Seljuk state of Rum had long since succumbed to the
Mongols and functionally vanished from the scene.
In the reign of Ala ad-din Kayqubad I (1220–1237), the Seljuks of Anatolia
were at their height. They had captured major port cities—Antalya and Alanya
on the Mediterranean and Sinope on the Black Sea (which they surrounded
with powerful fortifications), established a base in the Crimea, developed a
network of roads and fortified caravanserais in Asia Minor, expanded their
domain to Erzurum far in the east, advanced into upper Mesopotamia, and
established cordial relations with the Byzantine empire of Nicaea. There were
no major problems on the Byzantine eastern/Seljuk western frontier.
Meanwhile, beyond the eastern horizon, a far greater power than ever
imagined was drawing closer to the Near East. By the time he died in 1227,
Genghis Khan had gained control of a vast domain that stretched from the
Volga to Korea and embraced all of central Asia and eastern Iran. Temporarily
held back by succession disputes and the ephemeral Khwarezmian empire,
a Mongol embassy only arrived at the Seljuk court in 1236, demanding sub-
mission and tribute. The death of Kaykubad in 1237 and Mongol preoccu-
pation elsewhere postponed any serious action by either side until 1242
when massive Mongol armies moved on Mesopotamia and Anatolia. One
battle decided the fate of Rum: at Köse Dağ in the region of Erzincan on
June 26, 1243 the Seljuk army was annihilated, the sultan fled to Ankara,
and the Mongols advanced as far as Kayseri before returning to the steppes
of northern Iran. The world of Anatolia and its neighbors was changed
DOI: 10.1093/oso/9780198865438.003.0006
forever, especially in the west where Mongol advances displaced Turcoman

tribes, sending them fleeing toward the Byzantine frontiers.
The Seljuk vezir promptly headed for Mongol headquarters, where he
negotiated a peace that involved paying a heavy tribute, confirmed by the
appointment of the Sultan Ghiyath ad-Din Kaykhusraw II as the Mongols’
representative in Rum. The Seljuk state seemed to have escaped disaster as it
retained considerable autonomy. However, ominous troubles soon broke
out with a widespread revolt of people who were destined to be the foes of
settled government, the Turcoman tribes of the mountains of southern
Anatolia. To make matters worse, Kaykhusraw died in 1246, leaving three
sons: Izz ed-Din (age 11), Rukneddin (age 9), and Alaeddin (age 7). For
the next decade the history of the Sultanate is dominated by the struggles
of the factions behind these princes who never hesitated to call in the
Mongols for support. Izz ed-Din came to head up the pro-Byzantine fac-
tion, Rukneddin the pro-Mongol. At the same time, high officials and
aristocrats established their own power bases by taking over state lands to
secure their own incomes.
The next crisis came in 1256 when the Great Khan sent his brother
Hulagu to govern Iran, until then dependent on the Mongol horde of south-
ern Russia. When he arrived with his flocks and herds, Hulagu displaced
the Mongol forces based there, pushing them westward into Asia Minor.
There, they came into conflict with Izz ed-Din, who at that moment was
supreme. Despite Byzantine support, Izz ed-Din, defeated, fled into exile in
Byzantium. He returned to Konya in 1257 when Hulagu divided Anatolia,
the west (from Kayseri to Antalya) going to Izz ed-Din, the east to Rukneddin.
Such an action confirmed the powerlessness of the Seljuk government,
completely subordinate to the Mongol authorities. The following year, Hulagu
sacked Baghdad, abolished the Abbasid caliphate, and established the
Ilkhanate based in Iran. Anatolia had winter and summer headquarters for its
vast encampments scattered throughout the central plateau; the westernmost
winter pastures were around Akşehir and the plain of the Sangarius west of
Ankara.2 The previous occupants of these lands, including many Turcomans,
were pushed out, with most of them going to the frontier regions.
In 1261, suspected of intriguing with Mamlukes and Turcomans, Izz ed-Din
took refuge in Constantinople where he found a very changed situation.
Michael Palaeologus, recognizing recent developments, abandoned the
traditional alliance with the Seljuks in favor of the Mongols. He also formed
alliances with the Mamlukes and the Golden Horde. By these means,
Palaeologus protected his eastern frontier at a time when he faced major
The Overlords 159
threats from the west; he was safe from the Mongols, but not from the tribes.
As for Izz ed-Din, after spending a slothful time in Constantinople, he
wound up in prison from which he was rescued by Nogay who established
him in the Crimea, where he died in 1280.
For the next fifteen years, Anatolia, administered by the pervane
Muìneddin and the vezir Fakhreddin, enjoyed relative stability.3 When
Rukneddin tried to assert himself, he was killed, the nominal sovereignty
passing to his small child Kaykhusraw III (1265–1282). The beginning of
this period, though, brought a development that was to have drastic conse-
quences: the establishment of autonomous Turcoman principalities. The
first was in the mountains of southern Anatolia, where Karaman bey, head
of a tribe long established in the area, gained control of the region of
Larende and Ermenek around 1260 and used it as a base for launching an
attack on Konya two years later. This was driven back by the pervane with
Mongol support, but from then on, the Karamanids were to be one of the
most intransigent and persistent enemies of the Ilkhans. The same years saw
the rise of another Turcoman state based in Denizli, where a first attempt at
independence was squelched by the Mongols who, however, allowed the
local chief to maintain control of his region in 1262. The Mongols were
keeping control of the Anatolian plateau, but allowing the Turcomans to
establish themselves in the periphery.
In addition to providing pasture lands to the Mongols, the Sultanate had
to pay a substantial tribute in cash, textiles, and animals, often resulting in
large debts for which the Mongols appointed officials to secure payment. At
the same time, high officials were carving out domains of their own: one of
the most important was that of the sons of Fakhreddin around Kütahya and
Karahisar in Phrygia. Such developments seriously weakened the Seljuk
administration.
The events of 1277 brought Ilkhanid Asia Minor close to disaster. Early in
the year, Baybars, Mamluke Sultan of Egypt, renowned for his victories over
the Crusaders and for his participation in the battle of Ayn Jalut (1260),
which stayed the Mongol advance into Syria, moved on to Asia Minor. After
defeating the local Mongol and Seljuk forces at Elbistan in April, he occu-
pied Kayseri, where he was enthroned and named in the hutbe. In the next
month, however, he withdrew, fearing the approach of the main Mongol
army. His death in June ended this danger but by then a major threat had
arisen in the west. The Turcomans of the southwest, led by Karaman and
reinforced by Eşref and Menteşe, taking advantage of Baybars’ attack occu-
pied Konya in May after defeating Germiyan and the sons of Fakhreddin,
who had remained loyal to the Sultanate. The rebels proclaimed a man,
Jimri, who they claimed was Siyavush, son of the former Sultan Izz ed-Din.
He assumed royal power, issued coins in his own name, and appointed as
his vezir the Karamanid ruler Mehmet Bey, who ordered that the govern-
ment decrees should be in Turkish rather than Persian. In October, the
Ilkhan sent his brother with a large army that took Karaman, and defeated
and killed Mehmet Bey. Jimri, however, escaped. The following spring,
Fakhreddin and Germiyan took up the fight and managed to capture and
execute Jimri, bringing the rebellion to an end, but it had consequences.
Since the Turcomans of Burghulu and Denizli had not participated in
crushing Jimri, the Mongol force moved on those centers as well as Sandıklı,
Juhud, and Karahisar, where they installed the grandsons of Fakhreddin in
what became a separate emirate. The emirs who submitted suffered no fur-
ther punishment.
These events provoked a major change. So far, the Seljuk Sultanate was a
tribute-paying dependency of the Mongols, retaining a considerable degree
of autonomy. Now, the Ilkhanids stepped in and established direct rule.
Mongol governors became supreme as the Seljuks were reduced to the role
of puppets, while the finances were reorganized and centralized. In the new
regime, the civil administration was subordinated to the military com-
mander of Anatolia. But the coins still named only the Seljuk ruler; the
Mongols did not issue coins in their own name in Anatolia until the reign of
Ghazan, beginning in 1298.
In 1280, when the former Sultan Izz ed-Din died in the Crimea, his son
Mesud embarked for Asia Minor in the hope of regaining his rightful
throne. He came to terms with the ruler of Kastamonu and with the Ilkhan
Abaqa, who did not restore him to the Sultanate but sent him against
Karaman, where he made no move. Instead, his cousin Kaikhusraw III—
with Mongol support—defeated the Karamanid threat in 1282, only to be
executed two years later when a new Ilkhan had installed Mesud as Sultan.
Then yet another change of Ilkhans divided the Sultanate, giving the east to
Mesud and the west to the young sons of Kaikhusraw. Mesud, however,
soon took Konya, killed the princes, and gained undivided authority by
1286. The following years were complicated by the ongoing struggle with
Karaman and a new enmity with Germiyan, leading to an inconclusive
peace in 1288. All this was accompanied by growing Mongol control and
interference in the regular operation of government. The finances were
especially disrupted as the Ilkhans turned them over to officials sent from
Iran, imposed new taxes, and finally set up a system of tax farming, which
led to corruption and widespread discontent.
The Overlords 161
Gaikhatu, who had been governor of Anatolia and was well aware of
the perennial instability of the frontier region, became Ilkhan in 1291.
Determined to restore order, he arrived in Kayseri with a huge army in
October 1291. He devastated the lands of Karaman and Eshref, enslaving
thousands, then moved on to Denizli, whose resistance only led to its popu-
lation being massacred. From there, he ravaged the lands of Menteşe and
inflicted further devastation on Karaman before returning to Iran in June
1292. At the same time, a major revolt had broken out in Paphlagonia which
proved harder to control, being suppressed only in 1293. These years saw
deteriorating conditions as Gaikhatu turned out to be an incompetent
drunkard. Asia Minor only returned to some kind of order with the acces-
sion of Ghazan in 1295, who had to spend his first years suppressing revolts
that devastated Anatolia until 1299.4
Notes
1. The present sketch makes no pretense of originality, but is intended simply to

provide some background for what follows. It is based on Melville 2009, Sümer
1969, Cahen 1968, Togan 1970, Spuler 1955, and Kolbas 2011, 13–22.
2. See Smith 1999.
3. Pervane: a sort of right-hand man of the sultan: see Cahen 1968, 221f.
4. This narrative will be continued in the next chapter.
6
Osman and his Neighbors
Pachymeres, in his account of the devastation and loss of Byzantine Asia

Minor, names the culprits, defining these Turcoman forces by their leaders.1
Some were already well-established:
Alisurai (Alishir of Germiyan); Mantakhiai (Menteşe); Atmanes (the fol-

lowers of ‘Atman’ i.e., Osman);
some were new on the scene:
Atinai (Aydın); Ameramanai (probably the Candaroğlu); Amourioi; Pagdinai

(maybe Karesi)
others have not been identified, perhaps later absorbed into other groupings:
Alaides, Lamises, Sfondylai.
To these can be added Sasan, who appears in 1304, and the unidentified
Atares in 1305.
Note that the historian does not associate any of these groups with specific
areas: evidently they were not yet settled in the regions they came to domi-
nate or—in the case of those that already had a history—were emerging
from areas that were beyond the current horizon of Byzantium. This would
be the case of Germiyan in Phrygian Kütahya, Osman in interior Bithynia
and Menteşe whose Carian lands had been lost by Byzantium a generation
earlier.
Some were on the move to lands with which they would be identified.
The Aydınoğulları first appeared in Ionia in 1305 from a homeland under
the direct control or in proximity to Germiyan whose dependents they were
when they reached the Aegean region. The Candaroğulları, installed by
the Mongols in western Paphlagonia were, like Karesi, of obscure origins.
The Amourioi moved around between the Sangarius and Paphlagonia but
had no settled home, hence their request for Mesonesion.2 The followers of
DOI: 10.1093/oso/9780198865438.003.0007
Alais, of equally unknown origin were in a similar situation, for in 1303 they
arranged to share the citadel of Sardis with its Byzantine defenders. Atares,
who took Kouboukleia in 1305, seems to have faced the same problem.
Sasan had a base, but had not yet appeared on the scene as Pachymeres
describes it.
In these years (1302–1305) several groups were operating far from what
were or became their bases, as if the whole land were in turmoil: Osman
came from inner Bithynia to Nicomedia and the Sangarius; Turks from the
Maeander and Paphlagonia were with Osman at Bapheus; Sasa from Ephesus
to Philadelphia; Menteşe from coastal Caria to Tralles. In all this, tribal loy-
alties were flexible and alliances frequent. This would have been especially
true of Osman’s allies who came from the two regions which had huge
encampments of Turcoman nomads, ready to fight or loot, not yet organ
ized into states.3 As John Cantacuzene remarked: “it is the custom among
these barbarians, when one of them goes on a campaign, that those of
another satrapy who want to join it, are not pushed aside but received with
pleasure as allies.”4
These nomadic groups all needed a base to store food for the winter, as a
shelter for women and children, and especially as a place to keep the loot
they gained from one of their prime activities, predation. Alais and his
followers were quite clear about that need, for which an easily defensible
steep fortress would be ideal. Menteşe had solved this problem early on
by using a fortified island in a lake for what was apparently a considerable
treasure. It was hardly a coincidence that coins were struck there (in the
mint called Bafa) as early as 1298.
The parallel example of the Ottomans is typically cloaked in an edifying
folk tale about the lord of Bilecik agreeing to store their goods as they
moved between summer and winter pastures. Viewed more simply, this
could mean that Osman captured Bilecik, difficult of access and therefore
well suited to this purpose.
The tribes soon settled in the lands they had overrun, a situation
described by Nicephorus Gregoras (VII.214f), writing some forty years
later about the events of 1302. He believed that the Turks had agreed to
divide up the imperial lands in Asia Minor:
“Karmanos Alisyrios got most of inner Phrygia as far as Philadelphia

and the places closest to them up to Antioch on the Maeander from
there to Smyrna and the coast of Ionia another, called Sarkhanes;
Osman and his Neighbors 165
another ‘satrap’ called Sasan had already taken the region around
Magnesia, Priene and Ephesus one called Kalames and his son Karases:
from Lydia and Aeolis as far as Mysia on the Hellespont the land around
Olympus and all Bithynia next to it: another called Atman the land from
the Sangarius to Paphlagonia the sons of Amourios divided among
themselves.”5
This, then, was the age of settlement when the tribes could conquer new
and rich lands in the general collapse of imperial defenses. Settlement was
facilitated by the occupation of important urban sites, where the Turks
could find or install civil administrators and religious functionaries, secure
behind massive ramparts. The coastal emirates also got ports, bases of trade
for a new source of wealth. Menteşe acquired Milas and Palatia/Miletus;
Aydın got Ayasuluk/Ephesus. Karesi and Saruhan took major commercial
centers Pergamum and Magnesia; while the Ottomans had their eye (and
siege techniques) on Nicaea and Bursa.
Once established, the emirates that controlled parts of the Aegean
coast are striking for their power and prosperity. They rapidly became
players on the international scene, along with the Byzantine empire, the
rival Italian commercial states of Venice (represented by its governor in
Crete) and Genoa, the Hospitaller Knights of Rhodes, and the Frankish
principalities of Greece. These and the other emirates will provide a
context for Osman and enable him to be evaluated in comparison with
the rest. A survey of these states will also show the varying degrees of
information available about them. Thanks to surviving documentation,
Menteşe in Caria and Aydın in Ionia are the best known.6 Osman rises
in relative obscurity.
Osman
There is no doubt that Osman is a historical figure who laid the foundations
for empire by conquering territory from Byzantium and expanding his
small and remote domain to the thresholds of the three great cities of
Bithynia. Yet his lands were still smaller and poorer than those of some of
his neighbors.
Osman’s immediate neighbors were Kastamonu in the east, Germiyan on
the south, Karesi on the west and a few small and obscure emirates, also on
the east. Looming behind them was the power of the Mongols, the Ilkhanids
who ruled Iran and Asia Minor. The maritime emirates of the Aegean—
Saruhan, Aydın, and Menteşe—though not contiguous to Osman’s lands,
may be considered here to round out the image of western Anatolia. First, it
would be desirable to define Osman’s territory and to establish a chronology
for its expansion.
Osman’s career, as presented by the tradition, begins with mixed relations
with his Christian neighbors.7 The ruler of Inegöl, who harassed his annual
movements, was his enemy, but the tekfur of Bilecik, who stored Osman’s
heavy goods while he was on the move, was friendly. His first victory, dated
to 1285, was a raid that involved the capture and destruction of the small
fort of Kulaca, only two kilometers from Inegöl.8 The next year he defeated
a coalition of tekfurs at the Domaniç pass, but now the tradition becomes
problematic, having Osman directing his forces to the (Muslim) south, sup-
posedly conquering Karacaşehir in 1288 and sending a raiding expedition
into eastern Bithynia soon after.
Aşıkpaşazade (APZ) presents Osman’s northern frontier as remaining
surprisingly static until 1299 when, after avenging the treachery of the tek-
fur of Bilecik, he conquered Inegöl and Yarhisar. This brought him closer to
the coastal plains and their heavily fortified cities. Not long after, he took
Köprühisar and established Yenişehir which he made his capital and base
for moving against Iznik. In 1302, he defeated the coalition of tekfurs at
Dimboz and advanced as far as Ulubat, leaving Bursa increasingly isolated
and vulnerable, subject now to blockade by Osman.
So far, the tradition, portraying Osman as moving inevitably northward,
though at a very slow pace, interrupted by dealings with Karacahisar. But
now other sources help to illuminate the scene, as Pachymeres describes
Osman’s first victory over Byzantium at Bapheus in 1302. In a remarkable
coincidence Pachymeres and APZ both show Osman advancing up the
Sangarius where he took the strategic fortress of Malagina/Akhisar in 1304.
Likewise, Greek and Turk alike show him blockading and attacking Nicaea/
Iznik with increasing violence. Pachymeres alone explains what happened
next: the massive Mongol attack that defeated Osman and drove him from
all his recent conquests. The tradition is (not surprisingly) silent on this
momentous event, but like the Byzantine, has a hiatus of almost twenty
years. When the Ottoman advance resumes with the conquest of Bursa,
Osman is no longer the leader.
The tradition seems plausible enough in outline, however confused or
obscure the details may be: Osman’s attentions were directed to the north, to
the lands in the increasingly feeble hands of Byzantium, not only against its
three great cities—Nicaea, Nicomedia, and Prousa—but along the Sangarius.
The emirate of Osman saw the greatest changes of any of these states,
from an obscure hill country to the threshold of the great cities of maritime
Bithynia. Osman, according to the tradition, began with the territory inher-
ited from his father. This is worth considering to illustrate the beginning of
his career as well as some of the manifold problems of the sources. The
discussion will touch on the most famous and problematic of Osman’s con-
quests, Karacahisar, and consider the status of Eskişehir/Sultanönü, Inönü
and the cities and emirates of eastern Bithynia, under attack by Osman.
According to the tradition (APZ cap. 2), the Seljuk sultan Ala ed-Din set-
tled Ertuğrul on the frontier between Bilecik and Karacahisar whose tekfur
paid tribute to the sultan. Neşri, though, mentions the beys of Sultanönü,
Eskişehir, and Inönü and states that when Osman came to power, deputies
of the Seljuk Ala ed-din II were ruling in Sultanöyüğü, Eskişehir, and Inönü,
while the contemporary Haci Bektaş mentions Ermeni Beli as marking the
frontier between the Byzantine and Seljuk realms.9 This, then, was a frontier
area, with the main settlements in Seljuk hands. Ertuğrul seems to have
been content with the pastures around Söğüt assigned to him and not to
have embarked on any conquests.10
Ertuğrul faced an overwhelmingly powerful neighbor to the southeast
where the major city of Sultanyuki/Eskişehir was under Mongol rule, as
attested by the inscription (1266) and testament (1272) of Jibrail b. Jaja. The
Mongol policy of maintaining the image but not the reality of Seljuk rule
could account for Neşri’s statement about the deputies of the sultan ruling
the three cities. Sources reveal nothing of the relations between Ertuğrul
and the Mongols, but they certainly would not have been on the basis of
equality. It is quite likely that Ertuğrul’s function was to defend this section
of the Mongol/Byzantine frontier. In any case, Osman would have grown up
under the shadow of the mighty Ilkhanids.
Osman inherited a district around Söğüt, bounded on the north and west
by fortresses and villages ruled by Christians, evidently subject to Byzantium
whose historians make no mention of this remote and obscure district.
The closest and most important Byzantine outpost would have been
Bilecik, thirty kilometers northwest of Söğüt. Osman’s limits to the south and
southeast are more difficult to define because of ambiguities in the tradition.
Here, the major settlements were İnönü some thirty kilometers south,
Eskişehir about forty-five kilometers southeast and—most prominent in the
sources—Karacahisar ten kilometers west of Eskişehir. The history of these
places may help to understand the size and nature of Osman’s base and the
expansion of his lands.
İnönü, as already noted, finds no place in the accounts of Osman. APZ
only mentions the sanjak of İnönü being entrusted by Orhan to his son
Murad after the conquest of Bursa and before that of Iznik, with no indica-
tion of when and how it came into his hands (APZ 30). İnönü only appears
in history with the inscription of Hoca Yadigar of 1374 which indicates that
the place was independent or perhaps subject to Germiyan, implying that
neither Osman nor Orhan ruled it.
The case of Eskişehir is more complicated, partly because it had two
names: Eskişehir for the settlement around the hot springs and Sultanönü
or Sultanöyüğü for the fortress two kilometers away—but there was really
only one town, to which either name could apply.11 Neşri (I.73) treated the
two as separate cities. The sources about the Mongol governor Jibrail ibn
Jaja leave no doubt that the “Sultanyuki” and surrounding territory were
under Mongol control in the 1260’s. Half a century later, Timurtash’s rapid
conquests included Sultanönü which possessed no fortified cities but vast
plains—in other words it was independent in 1326, around the time of
Osman’s conquest of Bursa12 According to Ahmedi, the earliest but very
sketchy Ottoman source, Sultanönü was only taken by Murat I, together with
Ankara, in 1361.13 Its fate in Osman’s time is not clear, nor is its absence from
Balban’s list of the Turcoman emirates. Did it slip from Mongol control at
some point? Given the overwhelming power of the Ilkhanids, that seems
unlikely, but perhaps it had revolted and therefore had to be reconquered by
Timurtash. This entire region was necessarily affected by the settlement of
Germiyan in 1277, which interposed a powerful emirate between Sultanyuki
and the centers of Mongol power on the Anatolian plateau. But Germiyan,
however powerful, was still subject to the Mongols who summoned its rulers
and those of several other emirates to submit formally in 1316.
The appearance of Sultanyuki/Eskişehir in the tradition, which never
claims that Osman actually ruled the place, does nothing to clarify the situ-
ation.14 It seems most probable that the city was the center of a small emir-
ate. On the other hand, the evidence of pious endowments offers a different
point of view, for they include lands (admittedly on a small scale) near
Eskişehir and on the route there from Söğüt.15 If these really were endow-
ments of Osman, his lands would have reached close to Eskişehir.
Karacahisar, which looms great in the tradition, poses the greatest prob-
lems.16 Osman supposedly captured it in 687 (1288/9) when, according to
Neşri (I.87), he was thirty-five years old. The previous year the tekfur of
Karacahisar had come to the aid of the ruler of Inegöl against Osman who
defeated their coalition and, with the help of the sultan Ala ed-Din (who was
called away by a Mongol attack before the siege was completed), conquered
Karacahisar. He captured the tekfur, plundered the city and collected rich
booty. He distributed houses to his followers and others and thus made the
place an Islamic city (APZ caps. 5–6). Subsequently, it was his base for an
expedition to Yenice and Mudurnu (APZ cap. 10).
Delighted by the news of the capture of Karacahisar, the sultan is supposed
to have sent Osman important symbols of authority including a banner, a
tent, horses, and weapons (APZ cap.8). Another source gives a hint that
these honors were not altogether apocryphal, for the Mekece endowment
mentions Osman with the laqab or honorary name Fakhr al-Din.17 Since
these titles were normally conferred by a higher authority, such as the Seljuk
sultan, its existence may provide some confirmation for the sultan’s award of
honors, though without revealing the circumstances.
The most momentous event in the history of Karacahisar supposedly
happened in 1299 (APZ cap. 14): when Osman first occupied the place, it
was deserted. People from Germiyan and others asked for houses, which
Osman granted them. He held a market, converted churches into mosques
and named a kadi, thus proclaiming his independence from the sultan. This
date has been taken to mark the beginning of the Ottoman Empire. Settled
in his new conquest, Osman received a man from Germiyan who wanted to
buy the taxes of the market (APZ cap. 15). The tradition portrays the sim-
ple, noble Osman as being totally ignorant of such matters.
At this time, Anatolia was in turmoil from the revolt of the Mongol gov-
ernor Sulemis whose forces occupied most of the country from November
1298 to April 1299. He found a following among the Turcomans of the fron-
tier and it is possible that Osman took his side. It has even been suggested
that whatever rewards and honors Osman is supposed to have had from the
sultan, he actually had from Sulemis.18
By this time, Osman had made substantial conquests to the north includ-
ing Yenişehir, which he made his capital. He assigned the conquests to his
various relatives and captains (APZ cap. 16): his eldest son Orhan received
the sanjak (province) of Karacahisar, evidently considered a place of prime
importance. In 1304, when his ally Köse Mihal converted to Islam, Osman
made him Orhan’s partner in ruling Karacahisar (APZ cap. 20). Soon after,
while Osman was attacking Lefke, a body of Çavdar Tatars attacked the
market of Karacahisar. Defeated and captured by Orhan, they were par-
doned by Osman who wanted to live in peace with his neighbors. Finally
(APZ cap. 34), after the conquest of Iznik in 1331, Orhan granted Karacahisar
to his (otherwise unknown) cousin Gündüz; not surprisingly, the place had
fallen in importance after the greater conquests in the north.
Geography poses the greatest obstacle to understanding these accounts,
for Karacahisar is only ten kilometers west of Eskişehir, on one of the main
routes between that place and Kütahya, the capital of the powerful
Germiyanids. From whom, then, did Osman, who was still (in 1288) not
much more than a tribal leader, conquer it, and why was it so important in
1299 when his realm stretched as far as Yenişehir? Its proximity to a state far
more powerful than his own and the presence or influence of the Mongols
in Eskişehir make this a most unlikely place to have fallen to the insignifi-
cant Osman. The tradition associates Germiyan with this place in suspicious
narratives—that they requested houses in the newly conquered city and that
they knew about and wanted to collect market taxes—all suggesting an
important Germiyanid presence here or in the neighborhood. It seems saf-
est to conclude that the accounts of Karacahisar are all apocryphal, created
to make Osman the equal or superior to Germiyan and even the Mongols
who only appear here in the attenuated form of the Çavdar Tatars.
The evidence for these three cities helps to establish the limits of Osman’s
power. He had inherited the grazing lands of Ertuğrul, which stretched
twenty kilometers or less from Söğüt. Christian tekfurs ruled the lands to
his north and west, while Muslim states formed a barrier to the south and
southwest. İnönü was never his, nor apparently was Eskişehir/Sultanönü.
He may have acquired Karacahisar and maintained it as a fortified outpost
on his rear as he moved north, but even that seems unlikely. North was the
only direction available to him—and the only one where he could expand at
the expense of infidels rather than fellow Muslims. At this stage, he had no
possibility of moving against Germiyan or the far more powerful Mongols.
He had no major cities and had no special natural resources, but his lands
did lay astride a major trade route. It was from this unpromising beginning
that Osman embarked on conquests that were to produce a mighty empire.
Like other goals of Osman, the fate of eastern Bithynia poses problems.
APZ (10) reports that soon after the conquest of Karacahisar, Osman pro-
posed an expedition against Tarakçı Yenicesi. Mihal advised him of the route
and told him that the rich land of Mudurnu would be easy to attack because
Samsa Çavuş and his followers were settled near there (as they had been since
the days of Ertuğrul). Samsa joined them after they crossed the Sangarius and
together they raided Göynük and Tarakçı Yenicesi and reached Gölkalanoz
(Gölpazar), whence they returned to Karacaşehir. They took much loot but
no captives because Osman wanted to make the population his subjects.

Subsequently, after the successful Sangarius campaign of 1304, Konur Alp
took Düzpazar (Düzce) and Akyazi, defeated the enemy at Uzunca Bel, and
advanced as far as Bolu (APZ 22). This was apparently a raid.
After the conquest of Bursa, APZ (25) reports that Konur Alp took Akyazı,
Mudurnu, and Bolu as well as the land of Konurapa. The Anonymous (p. 5),
though, attributes the conquest of Tarakçı Yenicesi, Göynük and Mudurnu to
Orhan’s son Suleyman, as does APZ (34) who, however, places this after the
conquest of Iznik (or Izmit).19 He adds that these places had surrendered
willingly. Meanwhile, on the death of Konur Alp, Orhan had assigned his
lands to Suleyman (APZ 30)
Al-Uryan writing around 733, however, reports the existence of an emir-
ate called Koynuk Hisar.20 When Ibn Battuta passed through, the city,
inhabited only by Greeks, was Ottoman.21 It seems probable, then, that the
three towns named together by APZ were all part of an independent state
adjacent to Ottoman territory and only taken by Orhan. Osman’s lands did
not include them, though they were suitable for raiding.
It would be desirable to follow the chronology of these conquests.22 APZ
gives several dates (ed. Öztürk 337–9), which are generally dismissed as
inaccurate or fanciful:
684 (1285/6): Conquest of Kulaca: APZ 5

685 (1286/7): campaign of Ikizce: APZ 5
687 (1288/9): Conquest of Karacahisar: APZ 6.
699 (1299/1300): Conquest of Bilecik Yarhisar Inegöl: APZ 12–13; Osman
named in Friday prayers at Karacahisar: APZ 14
702 (1302): Victory of Dimboz: APZ 17
704 (1304): Conquests along the Sangarius including Akhisar: APZ 20
705 (1305); Further conquests along the Sangarius; beginning of blockade
of Iznik: APZ 22
726 (1326): Conquest of Bursa: APZ 23
731 (1331): Conquest of Iznik: APZ 32
735 (1335): Conquest of Karesi: APZ 35
Of these ten, the first five cannot be checked with another source, but the
date of the conquest of Akhisar finds confirmation in Pachymeres who
mentions the fall of Malagina in 1304, while Osman’s approach toward Iznik
of the following year corresponds with Pachymeres’ chronology. Dates for

the conquests of Bursa and Iznik are correct, only the acquisition of Karesi
is far off the mark. Considering that all these dates are not fanciful and that
several are accurate, it might be prudent to accept the earlier as giving at
least a possible chronological structure to Osman’s progress to the north.23
Karacahisar, as usual, poses insoluble problems.
The epic narrative of APZ presents Osman rising from the obscurity of
Söğüt to the conquest of inner Bithynia. Supplemented by Pachymeres, who
shows him defeating a Byzantine force at Bapheus, counts him as one of the
leaders who overwhelmed western Asia Minor, and leaves him in defeat
after a determined siege of Nicaea, the story keeps Osman in the back-
woods. When the Ottoman forces reappear after a long (and unexplained)
hiatus, Orhan is in charge, still relatively poor, operating on a small scale
and trying to capture a major city. There are no coins that can be attributed
to Osman with any degree of certainty, nor any buildings. In addition to
the sources already presented, though, surviving documents offer a more
nuanced picture.
The foundation document of Mekece issued by Orhan in March 1324
implies that Osman (who presumably died not long before) had brought
security to the Sangarius region and had it firmly under control. Osman had
conquered the crucial fortress of Malagina in 1304.24 He most probably lost
it to the Mongols in 1307, but evidently regained the region before 1324.
The style and language of the document indicate that the Ottomans by then
had developed a sophisticated political structure. Since it was issued before
the conquest of Bursa, it was probably drawn up at Yenişehir.
The evidence of this document may be supplemented by the grant by
Osman of a property in Muttalip north of Eskişehir in a region under
Mongol control as late as 1272. This raises a central question: what were the
relations between Osman and the Mongols?—a delicate subject skirted by
APZ but fundamental to understanding the rise of the Ottomans. In this
case, did Osman acquire his property peacefully under the aegis of the
Mongols, or did he conquer it from them? Given the disparity of forces at
this time, the former seems far more likely.
Germiyan
The most powerful of the emirates, has the longest history. Germiyan,
apparently the name of a tribe not an individual, first appears in 1239, based
around Malatya in southeastern Turkey and subject to the Seljuks; it is

mentioned again in 1261 when Kerimuddin ibn Alishir-i Germiyani was at
the Seljuk court.25 By 1277 the Germiyan had settled at Kütahya, apparently
awarded them as a land grant by the Seljuk regime for their help in sup-
pressing the Jimri rebellion. In 1286, however, they broke into revolt, devas-
tating neighboring regions until a massive Mongol army restored control in
1291. Their long conflict with the ever-weakening Seljuk regime, however,
was evidently successful, for an inscription of 699/1300 commemorates
repairs to a mosque in Ankara by the great emir Yakub ibn Alishir in the
days of Keykubad.26
With their strategic location at the edge of the plateau and their power-
fully fortified capital Kütahya, the Germiyan were in a position to dominate
the western emirates like Menteşe and Aydın, recorded as subject or
dependent to them. The Destan introduces Mehmet Aydınoğlu and his sons
as opening the gate of conquest for Germiyan, holding the stirrup for
them.27 Confirmation of Aydın’s subordinate role comes from Eflaki
(c.1290–1360) who wrote biographies of the leading dervishes. Early in the
fourteenth century, Arif Chelebi, head of the order, met the son of Alishir
who had camped with his army outside Ladik; among the entourage of the
Germiyan ruler was the son- in-
law of Mehmet Aydınoğlu, whom he
described as a subashi (military commander) of the son of Alishir. When
Arif visited Birgi, he met Mehmet, also described as a subashi of Germiyan,
who had not yet conquered the region.28
When Chelebi visited the camp, the son of Alishir made a very bad
impression by paying no attention to the religious devotions of Chelebi and
his disciples, for “he was a Turk and without ceremony and uninformed
about the world of the Friends of God” whereupon Chelebi rushed off, hurl-
ing insults at the Germiyan chief. An earthquake followed immediately.
Alishir’s son promptly sent apologies and gifts, and was forgiven. Later,
when Chelebi visited Kütahya, Yaqub Beg became his disciple.29
In 1304 when the Germiyan make their appearance in western Anatolia.
Pachymeres calls them “the Karmanoi around Alisyras, the most powerful
of the Persians” and consistently distinguishes the Karmanoi from their
leader Alisyras: Alisyras with the Karmanoi attack Philadelphia; Alisyras
takes Tripolis which he uses as a base, makes attacks together with the
Karmanoi.30 In each case, Pachymeres is actually referring to Yakub, son of
Alişir, who ruled from before 1300 to about 1340. Gregoras, on the other
hand, simply calls the leader Karmanos Alisyrios.31 Yet the consistent dis-
tinction between Alishir and Germiyan suggests a family enterprise, with
several (autonomous?) tribal units serving under one chief—or one ruling
family.32 The appellation given to Menteşe, “Karmanos Mantakhia” suggests
that he, too, was a relative or dependent of Germiyan.33 In any case, his son-
in-law Sasa was fighting in alliance with Germiyan against the Catalans at
Philadelphia before breaking away to become independent and then to call
in the Aydınoğullari.
Under Yakub, Germiyan rose rapidly; by 1304, they had captured for-
tresses near Philadelphia as well as the important stronghold of Tripolis on
the Maeander. By 1314, Philadelphia was paying tribute to them. Yet within
two years, the sons of Alişir had to face a powerful Mongol army, led by the
governor Choban who received the submission of Germiyan along with
Hamid, Eshref, Sahib Ata, and Kastamonu in his winter quarters.34 Germiyan
is here described as “the beys of Germiyan from Kütahya with the sons of
Alişir from the castles of the region,” maintaining the distinction between
Germiyan and Alişir found in Pachymeres. In the following decade, an
endowment provided for a zaviye in Uşak (1321) and in 1325 the sultan al-
Kermiyaniya built the castle of Sandıklı in Phrygia.35 They seem not to have
aimed at conquest in the west, but Yakub’s reign is poorly attested. In any
case, his capital Kütahya was the mint for a coin type of the Ilkhan sover-
eign Ghazan (1295–1304) struck in 698/1298 and 700/1300. Coins in
Yakub’s own name appear to belong to the later years of his reign or to
Yaqub II (1387–1399).
Osman and Germiyan
The tradition is consistently hostile to Germiyan whenever it appears in the

story of Osman:
APZ 2: Germiyan and Çavdar Tatars constantly raid (Afyon) Karahisar and
Bilecik; Ertuğrul resists them, gives security to the infidel population. When
he takes over, Osman follows the same policy. His favor to the non-
Muslims is matched by hostile relations with Germiyan that begin here.
APZ 6: Germiyan stirs up the infidels, allies with the tekfur of Karacahisar
against Osman.
APZ 9: Osman maintains good relations with the tekfur of Bilecik and with
Mihal, but is always hostile to Germiyan; the infidels are pleased by
this enmity.
APZ 9: a man from Germiyan refuses to pay an infidel merchant for a

drinking glass; is chastised by Osman who gains a favorable reputation
among the infidel.
APZ 12: mention of the hostility between Osman and Germiyan.
APZ 14: people from Germiyan come to Karacahisar, ask Osman for
empty houses
APZ 15: a man from Germiyan asks Osman for the right to collect taxes on
the local market.
APZ 21: Çavdar Tatars from Germiyan attack the market of Karacahisar
In all this, Ertuğrul and Osman are defenders of the local non-Muslim pop-
ulation against the Germiyan, who appear to be city dwellers and taxpayers.
When the deep- seated hostility turns to fighting, the Ottomans win.
Nothing suggests that the Germiyan are more than unpleasant neighbors, in
no way superior to the Ottomans.
The hostility at least is real enough, as in the account of Ibn Battuta who
is warned that the Germiyan are dangerous bandits and in the detailed
description of Balban who notes the hatred the other emirs felt toward
Germiyan.36 In their case, Germiyan’s overweening superiority stirs enmity,
but there is nothing in the tradition to reflect the real power of this neighbor.
The tradition presents only a distorted view, far too limited to enable
anything about Germiyan or their actual relation to the Ottomans to be
perceived. It only provides vague hints of a very different reality.
Kastamonu
The vast mountainous region that stretches northeast from Osman’s terri-
tory to Kastamonu and Sinope contained important trade routes but no
large cities. Its history is obscure and often confused. For much of the thir-
teenth century, Kastamonu and its region were under the domination of the
Çobanoğulları established by Husameddin Çoban bey, a successful Seljuk
commander, around 1225. During the reign of his successor Alp Yürek, the
region became subject to the Mongols.37
Marino Sanudo, describing the situation around 1275, blamed Michael
Palaeologus for failing to defend one of his most valuable and powerful
provinces, Paphlagonia. This large, well- watered and well cultivated
region was constantly infested by the Turks, especially by their leader,

the “crudelissimo Canino Caraman.” An Italian renegade, the grand duke
Licario, won a major victory, but the gains were all lost when troops were
recalled to the capital to fight the Latins, leaving the empire only with Pontic
Heraclea on the sea.38
Alp Yürek’s son Yavlak Arslan became involved in the competition for the
Seljuk sultanate and wound up on the losing side, being defeated and killed
by a joint Mongol-Seljuk force in 1292. The victors turned the district of
Eflani (in western Paphlagonia) over to their supporter Yaman Candar,
founder of a new dynasty, but he had hardly taken control when Yavlak
Arslan’s son Mahmud seized power and ruled until 1309, when he in turn
succumbed to Yaman’s son Suleyman whose dynasty, the Candaroğullari,
remained in power until 1461. Suleyman captured Burghlu (Safranbolu) and
Kastamonu from the Çobanoğulları and moved his capital to Kastamonu
when he took over the region. In 1316, he was one of the emirs summoned
to submit formally to the Mongol commander Ctoban, who had taken up
winter quarters in Karanbük.39 His principality became much richer, more
important, and more strategic when he took Sinope in 1324, allowing him
to participate in the rich trade across the Black Sea.
Sometime before 1313, a Turkish commander known only by his titles
Gazi Çelebi, took Sinope from the Genoese. He built up a powerful fleet
including eight galleys at a time when the beyliks had only light vessels, and
used it to attack the Crimea where the Genoese had an important colony.
His treacherous massacre of Genoese allies and his attack on a Venetian
fleet made him notorious among the Christian powers. When he died in
1324, his emirate came to an end and the city was incorporated into the
Candaroğlu realm.40
The figures of Amourios and his sons Ales Amourios and Nastratios,
narrated by Pachymeres, find no certain parallel in Turkish sources. They
were apparently chiefs who operated between the Sangarius and Kastamonu
in the years around 1300 and at one time were allied with the Candaroğlu.41
The emirate remained closely associated with the Ilkhanids until the
1330s.42 Coins struck in their capital Kastamonu were of Ilkhanid type,
bearing the name of the Mongol ruler; they become more regular and abun-
dant from 723/1323.43
The emirate of Kastamonu, then, was a neighbor of Osman influenced or
dominated by the Mongols. It also had a large population of Turcoman
tribes. The Arab geographer Ibn Saìd (1213–1286), reported that there
were 100,000 Turcoman tents in the neighborhood of Kastamonu and
another 30,000 at “Karabuli” west of Ankara, which may have been another
place in Paphlagonia. They provided horses, mules, and slaves.44 They were
also a source of instability, revolting against the Ilkhanids, notably in 1301 in a
massive operation that took the Mongols two years to subdue.45 They proba-
bly formed the Turks from Paphlagonia who joined Osman at Bapheus.
Four Minor States
Strictly speaking, Kastamonu was not Osman’s direct neighbor, for two
small emirates lay between them, based in Göynük, adjacent to Osman’s
conquests along the Sangarius, and Gerede, about halfway between Osman’s
lands and Kastamonu. Since they are only known in Orhan’s reign, they will
be discussed in the next chapter.
Adjacent to Osman’s southern frontier were the mini-states of Sultanönü
(Eskişehir) and İnönü, discussed above. Virtually nothing is known
about them.
An Anomalous City
Philadelphia first appears in this account in 1259, when the new emperor
Michael Palaeologus brought his main army there in a show of force. He
inspected and strengthened the frontier defenses, and made generous dona-
tives to the defenders. The city was of special importance because it com-
manded the easy route to Laodicea and the upper Maeander, a major base
of the Turcomans, and so was a frequent object of their attacks.46 Thus in
1280, Michael’s son Andronicus secured the region of Philadelphia, as did
Philanthropenus in his successful campaign of 1293. But geography could
not be defeated, and the city found itself under serious attack by Alişir of
Germiyan and his subordinates Sasa and Aydın in 1304. When the Catalans
arrived in the aftermath of the collapse of Byzantium’s Asiatic frontiers they
found the fortresses of the region, notable among them Tripolis on the
Maeander, conquered by the Turks. The Catalans were victorious in a battle
that rescued the city, and made it their base of operations. Late in 1304,
however, the Catalans were recalled by the emperor; they never returned to
Anatolia. As soon as they were gone, Germiyan resumed the attack (in
1305), reducing the Philadelphians to the misery of a siege.47 In all this, its
powerful fortifications enabled Philadelphia to survive.
In 1310, Philadelphia, already twice attacked by Germiyan, suffered the

torment of another siege, which was lifted in exchange for a heavy tribute.
The Turkish success was commemorated in the dedicatory inscription of a
medrese in Kütahya dated 1314 which states that it was built with the trib-
ute, jizya, of Alaşehir. This term was normally used to denote the tribute a
Muslim state imposed on its non-Muslim subjects.48
By this time, a major change had occurred in the nature of Philadelphia’s
strategic location, for the establishment of the emirate of Saruhan with its
capture of Magnesia in 1313 meant that Philadelphia was cut off from the
Aegean and from any direct connection with the Constantinople. Its loca-
tion gained a different significance for now it was at the crossroads of three
often rival emirates—Germiyan to the north, Aydin to the east and south
and Saruhan to the west. Rivalry between these states may have been a factor
in the city’s survival as a Christian enclave within an Islamic land.
Germiyan and Aydın returned to the attack in 1322, when they besieged
the city and its outlying bulwark, Fort St. Nicholas. Although they employed
catapults and made the inhabitants miserable for a year and seven months,
they failed to conquer. The emperor got the news that Philadelphia, besieged
by Turks of the immediate neighborhood (i.e., Germiyan), was on the point
of surrendering, but he was in a hopeless position: he had no weapons or
army capable of facing the enemy, nor could he help a city so far from the
coast, surrounded by many enemies. He called on Philanthropenus, blinded
and exiled for twenty-eight years, and sent him without troops, weapons, or
gold, armed only with his experience and understanding. While he was still
on way, the Turks abandoned their attack and came to meet Philanthropenus
as friends, remembering their good relations in the past; some had even had
military training by him. They promised to grant him every wish and as a
result Philadelphia was saved and soon reached the height of prosperity.
It drew its wealth from its fertile territory and successful trade in high-
quality textiles, especially silk, and leather.49 Its merchants are attested as far
away as Thessalonica and Selymbria.50 Although Philadelphia was now cut
off from imperial territory, the emperor was able to send Philanthropenos
(though not an army) and to appoint the governor of the city.51 In other
words, Philadelphia was still part of the Constantinople, though the means
by which this connection was maintained are unknown.
Soon after, in 1327, the ferocious Mongol governor of Anatolia, Timurtash,
arrived in Philadelphia, not to attack the city but to bring Germiyan and
Aydın to order and to support the independence of Philadelphia, subject
to the emperor who was allied with the Mongols.52 Mongol influence,
however, was nearing the end, for the Ilkhanid state fell apart after the death
of its last effective ruler, Abu Sa’id, in 1335.
Karesi
Kalames and his son Karases possessed an exceptionally large territory

stretching from Lydia to the Hellespont, that is the entire northwest of
Anatolia, with coastlands on the Aegean and the Sea of Marmara.53 The two
leaders claimed descent of the Danishmendids who had ruled eastern
Anatolia in the twelfth century.54 They apparently subdued what was to be
their homeland around the turn of the fourteenth century when Byzantine
fortunes were rapidly declining and it was probably to them that the port of
Assos and the inland fortress of Kenchreai fell in 1304.55 At the time of the
Catalan expedition, Achyraous and Chliara were still Byzantine, but “Turks”
had taken Germe.56 They were presumably the Turks who by that time had
gained control of the whole (northwestern) coast, with the exception of
Adramyttion and Phocaea, held by the Genoese.57
These vague references are all that can be taken to reflect the establishment
of Karesi in northwestern Turkey. There are no inscriptions or monumental
buildings, and but one coin type. This was struck in Bergama in the name of
Uljaytu (1304–1316).58 It would thus have been issued during the reign of
the founder of the dynasty, Karesi Beg (1297?–1327), in his capital.
Saruhan
Surprisingly, in view of the extensive rich lands of the Hermus valley that
they controlled, there seems to be no information available about Saruhan
at this time beyond the mention by Gregoras of their settlement in Lydia.
No coins, whether struck in the name of the local ruler or Seljuk or
Mongol overlords, are known from Saruhan in this period.59
Aydın
Aydın first appears in history in 1304, when its forces, together with Sasa,
were attacking Philadelphia, only to be driven off by the Catalans. They
were probably subordinate to Germiyan as they certainly were in the
following year when Sasa called Mehmet of Aydın and his sons into the
region of Ephesus. By 1307, Mehmet had dispossessed Sasa, and taken
control of the Cayster and Maeander valleys that were to be the core of his
family’s territories. He made his capital at Birgi on the mountain slopes
overlooking the central Cayster valley.60 It was there that he built the first
monumental mosque of any of these emirates. The inscription that
announced his conquest of Birgi in 1307 and the construction of the
mosque in 1312 reflects his independence.61
Aydın returned to the attack on Philadelphia in 1322, again in the company
of Germiyan, though whether as an ally or subordinate is unclear. They were
still operating together in 1327 when the two powers, again threatening
Philadelphia, were called to order by the Mongol governor Timurtash.62
Mehmet struck coins of which one type is known, bearing neither date
nor mint mark. Others were struck in the name of overlords: the Seljuk
sultan Mesud II, who reigned intermittently between 1282 and 1307, his
issue dated 1306 was struck in Ayasuluk (Ephesus); and the Ilkhan
Khudabende Uljaytu whose issues of 1310 were minted at Ayasuluk and
sultan Hisar.63
There is less information about the naval exploits of Aydın in this period,
when it pursued the same activities as Menteşe, often carrying out joint
raids with its neighbor against the lands held by the Christians. In 1317,
Mehmet of Aydın took the city and castle of Smyrna from the Byzantines
but still faced an obstacle in the form of the castle that commanded the
port, held by the Genoese.64 That was only to fall to Mehmet’s more famous
son Umur a decade later. Meanwhile, Aydın controlled the important port
of Ephesus from which their fleet, with 2600 men aboard, attacked Chios,
only to be defeated by the knights and Genoese in 1319.65
Menteşe
This emirate is one of the first to enter history—in 1277, when its forces
joined those of Karaman and Eshref in the revolt that led to putting the
pretender Cimri on the Seljuk throne in Konya. The revolt was forcefully
suppressed by the Mongol army two years later. These disturbances appar-
ently caused the Menteşe to move from their original home, Makri
(Fethiye), toward the Byzantine frontier.66
They first appear in a western source in the reign of Michael Palaeologus
(d.1282), when the “Turquenodomar Mandachia” invaded the lower
Maeander valley and in 1284 when “Salpakis Mantakhias,” evidently the

founder of the dynasty, conquered Tralles after its abortive restoration by
Andronicus III.67 In 1291, after another widespread Turcoman revolt, the
Ilkhan ruler Gayhatu led a devastating campaign which ravaged the lands of
Menteşe (their location not specified).68 Two years later, it was the turn of
Byzantium as the imperial general Philanthropenus captured the fort of
Dyo Bounoi and its treasure from Menteşe’s widow.69 This stronghold was
on a virtually impregnable island in the Milesian lake near the mouth of the
river, a safe place for Menteşe to store the great wealth he had accumulated.
So far, the Menteşe clan (the founder seems to have died around 1290) has
been associated with the Maeander and Cayster valleys and with the region
of Miletus, but their original base was apparently on the Carian coast, lost
by the Byzantines before 1269 and the natural home of piratical fleets.70
Menteşe took to the sea, its fleet perhaps augmented by sailors from the
Byzantine fleet disbanded in 1285 who are recorded to have become
pirates.71 It was probably discharged Greek sailors who came to form the
backbone of the fleets of the maritime emirates, for the Turks themselves
had no experience of the sea.72 Attacks on Rhodes in 1300 and Chios in
1306 by Turkish fleets may have been the work of Menteşe.73
On land, in 1304, the “gabella de Mondexia” attacked the Catalans at
Thyraia/Tire in the Cayster valley and the next year, after the Catalans had
left, the “Persarkhos” Sasan, son- in-
law and follower of “Karmanos
Mantakhias” took Ephesus, Pyrgion, and Thyraia.74 He had already, in com-
pany with Aydın and Germiyan, been attacking Philadelphia until driven
away by the Catalans. Sasa soon parted company with Menteşe, established
his own realm comprising Ephesus and the Cayster valley, and called in
Mehmet son of Aydın and his four sons, all clients or dependents of
Germiyan. Soon after, Sasa broke with them, joined the Christians, who had
mounted a naval expedition, and was killed in the fighting. The subsequent
history of his lands belongs to Aydın.
The name “Karmanos Mantakhias” raises a surprising possibility.
Pachymeres, as noted, is careful to distinguish the Germiyanid commander
Alishir from his followers, the Karmanoi. Gregoras, describing the same
events, refers to Alishir simply as “Karmanos Alisyrios.”75 “Karmanos
Mantakhias” seems to fit the same pattern, perhaps identifying Menteşe as a
follower or relative of Germiyan. This would suit the context, with Sasa as
the key figure—relative of Menteşe and ally of Aydın.76
When more information is available, Menteşe appears indulging its two
main activities: raiding and trading. In 1311, 250 merchants from Rhodes
visited Menteşe to buy provisions and animals; the next year 23 “Turkish”
ships, evidently of Menteşe, attacked Rhodes, only to be crushed by the
Knights of Rhodes.77 In 1320, Orkhan, Menteşe’s emir, sent out a much
larger fleet—over eighty ships—in a major attempt to subjugate Rhodes,
once again was defeated by the knights, this time supported by the Genoese.78
Despite these attacks, trade continued, and Menteşe started to grow rich
with the proceeds primarily from agricultural products and animals.
No monumental buildings nor inscriptions have survived from this
period—the reign of Mesud son of Menteşe, c.1300–1320)—but coins were
struck in Menteşe lands between 1298 and 1303 in the name of the Seljuk
sultans Kaykubad III (three mints) and Mesud II (four mints).79
The mountains of Caria and Lycia—the interior regions of Menteşe—
presented a very different situation. According to the Arab geographer Ibn
Saìd (1213–1286), the whole region from Antalya to Denizli was the home
of 200,000 tents of Turcomans who constantly raided the coastal settlements,
stealing children whom they sold to the Muslims.80 They also made rugs
which were exported through the port of Makri (Fethiye). Though their
numbers are greatly exaggerated, these nomads could be a menace to more
than their neighbors, by joining in fights even far away. They no doubt account
for the Turks from the Maeander who fought alongside Osman at the battle
of Bapheus.81 It is not surprising, then, that the Ilkhan Gaikhatu marked
Menteşe for severe devastation as punishment for the Turcomans of the region
who had revolted against the Ilkhanids. The relation between the nomadic
Turcomans of the interior and the settled outward-looking state of Menteşe
are unknown.
The Mongols
Ghazan, who was Ilkhan (1295–1304) when Osman appeared on the scene,
brought stability after the chaos of the preceding reigns.82 He first had to
secure Asia Minor where Baltu, the governor he sent in to restore order in
1296, himself revolted the following year. The rebellion was suppressed by
Sulemis who was made supreme military commander, but he in turn in
November 1298 revolted with Mamluk and Turcoman support. The large
force he commanded—reportedly of 50,000—took control of much of
Anatolia but was decisively beaten by Ghazan in April 1299. Sulemis
escaped to Mamluk territory, raised another army with help from Karaman,
Eshref, and the Mamluks, and advanced on Ankara where he was defeated
by Ghazan’s commanders Qutlughshah and emir Choban. His execution in

September 1299 finally brought some stability.
Close relations with Byzantium were restored following an embassy from
Andronicus II late in 1302 which requested help against the Turks and pro-
posed a marriage alliance.83 Ghazan accepted the terms and agreed to the
marriage with an illegitimate daughter of the emperor. Ghazan’s death in
1305, however, threatened this arrangement but his successor Uljaytu held
to the terms of the treaty by sending a large force to help the emperor
against Osman. The other Turcoman chiefs, fearing the Mongols, cut short
their attacks on Byzantium.
These years also saw the end of the Seljuk sultanate, itself long powerless,
providing only figureheads whose ambitions were inevitably squelched by
the Mongols. In 1297, Mesud II, involved in Baltu’s revolt, was deposed,
exiled to Tabriz, and replaced by his nephew Kaykubad III whose corrup-
tion and reputation for bloodthirsty vengeance on his enemies led to his
deposition and execution in 1303. Mesud was then restored once again to
the powerless sultanate. He was the inglorious last representative of the
Seljuks of Rum, dying in obscurity around 1308.
With Anatolia finally under control, Ghazan was able to embark on a
perennial Ilkhan ambition, the conquest of Syria. Beginning in November
1299, he managed to take Aleppo and Damascus, but could not hold them
and withdrew in January 1301. Another attempt in 1303 also failed.
Ghazan was a reformer who aimed at centralization, uniformity, and sta-
bility. His efforts are particularly notable in a comprehensive reform of the
coinage carried out in 697 (1298). This provided a standard coinage in sev-
eral distinctive denominations issued by mints throughout the empire.
Their uniform design changed as the coinage was gradually devalued, with
the weight of the dirham (the preferred denomination in Asia Minor)
declining from 2.28 g under Ghazan to 1.62 in the last years of Abu Saìd.
Emir Choban was a distinguished military commander who had fought
his first battle in 1289 and been continuously in the service of the Ilkhanids.
In 1316, following widespread disturbances on the death of Uljaytu, when
Karaman had taken Konya, Choban moved west against the Turcomans. He
took up winter quarters in Karanbük where he received the submission of
the inland emirates: Felekuddin Dundar of Hamid, the Eşrefoğulları of
Gurgurum, the sons of Sahip Fahruddin of (Afyon) Karahisar, the beys of
Germiyan and the sons of Alişir of Kütahya and neighboring castles,
Suleymanpaşa of Kastamonu as well as the Armenian ruler of Sis.84 These
were evidently the emirates that had common frontiers with the Ilkhanid
domain and represent the geographic zone of broken mountainous country

between the central plateau and the coastal plains. The maritime states and
the Ottomans were beyond Choban’s reach (or interest). The emirs all
brought suitable gifts and were sent back in peace to their states. This pow-
erful demonstration established a general stability with emirs of the frontier
directly subordinate to the Mongols, but Karaman had not appeared.
Consequently, the next year Choban crushed the Karamanids and success-
fully recaptured the old Seljuk capital.
After settling affairs in Anatolia, Choban went east to the court, leaving
his son Timurtash in change of Anatolia. Under the new ruler, Abu Saìd,
who was only twelve when he ascended the throne, Choban rapidly rose to
dominance. Named commander of the armies in 1317, he married the
Ilkhan’s sister and appointed his sons to govern the most important prov-
inces. After crushing a plot by rival emirs, he was functionally the ruler of
the empire. For a decade, his power was unshaken, but personal problems
led to his downfall, beginning in 1325 when he refused to allow Abu Said to
marry his daughter. From that moment, the Ilkhan was determined on his
destruction; he found a suitable occasion in 1327 when Choban’s overween-
ing son Dimashq Khwaja was accused of violating the Ilkhan’s harem. He
was put to death and order went out to destroy the entire family. Choban
fled to Herat where the local ruler had him executed.
Timurtash in the meantime was earning a ferocious and extravagant rep-
utation. In 1320, after Karaman again took Konya, he responded in force,
capturing the city and inflicting severe reprisals, much as he did against
Cilician Armenia in the next year. In 1323, he took prisoner his worst ene-
mies, the beys of Karaman and Hamid. Success evidently went to his head,
for he started to act independently, striking his own coins (which have not
survived), inserting his name in the hutbe and even claiming to be the
Mahdi. He was rescued by his father Choban who took him as prisoner
before Abu Sa’id who pardoned him and restored him to his post. His sub-
sequent conquests and disaster belong to the reign of Orhan.
Al-Umari, writing in the 1340s, described the relations between the
Ilkhanids and the Turcomans. He explained that when the Mongols took
over, the Seljuk sultans were left with titles but no authority. Real power was
in the hands of the Mongol governor; the Friday prayer was made in the
name of the princes descended from Genghis Khan; gold and silver coins
were struck in their name. When the Seljuks were in the last stages of col-
lapse (around 1300), the Turcomans took over a large part of western
Anatolia, particularly the mountainous regions where Mongol forces could
not operate. Nevertheless, they didn’t stop seeking the goodwill of the
Mongols to ensure possession of the regions they had occupied. This situa-
tion, marked by submission and revolt, good faith and bad, continued for a
long time.
Later, when the Ilkhanids were in decline (1320s and 1330s), the Turks
consolidated their power. They recognized the predominance of Germiyan,
but each was independent, with complete control over his territory. Their
constant occupation was war against their infidel neighbors. Since each was
jealous of the others, they sought the support of greater powers, the
Mamluks of Egypt and the Ilkhanids. Several sought formal appointments
as delegates of the Mamluk sultan, receiving signs of honor—standards,
banners robes of honor, swords, and horses. The Turcoman emirs, despite
the strength of their mountain position and armies didn’t stop courting the
goodwill of the rulers of the family of Genghis Khan, sending rich presents
and maintaining agents in the Ilkhan court. They all said the Friday prayer
in the name of the reigning member of the house of Hulagu, and in particu-
lar sought the favor of their neighbor, the Mongol governor.
A Bright Future for Some
An observer, contemplating the Turcoman states that had risen from the
ruins of Byzantium in western Asia Minor in the early fourteenth century,
would probably have concluded that the future belonged to the maritime
Menteşe or Aydın. Both were enormously rich from the trading and raiding
that produced monumental buildings in new capitals and was reflected in
the production and circulation of abundant silver coins. Each had large and
prosperous cities conquered from Byzantium and maintained as centers of
administration, trade, and production. Though less well known, Karesi’s
strategic location on the Aegean and the Marmora enabled it also to trade,
raid and provide mercenaries.
A closer look would reveal the emirate of Germiyan, which dominated
Aydın and Menteşe and was recognized as the most powerful of all these
states and appeared to have an unshakable position, no doubt helped by its
central location bordering Karesi, Saruhan, Aydın, Hamid, and the lands of
Osman and the Mongols. The future of the outlying Candaroğlu emirate
also looked bright after it acquired Sinope, center of the Black Sea trade.
Beside these flourishing states, the landlocked Ottoman emirate would have
seemed a minor player in a remote area, without large cities or important
natural resources. Militarily, though, it had potential, especially because it

directly faced the declining but still impressive Byzantine empire. Osman
had won a significant victory over a not very large Byzantine army, was
poised to capture the powerfully fortified Nicaea, and appears to have estab-
lished control over the route along the Sangarius giving his forces access to
Nicomedia and the Black Sea coast. But these conquests looked ephemeral
after 1307 when a massive Mongol army moved against him, driving his
forces from the plains and recent conquests back into the hills. Osman van-
ishes from the scene for almost twenty years; when he reappears with his
son Orhan, the situation is about to change drastically.
APPENDIX
The Date of the Death of Osman
This seemingly straightforward question turns out, as usual, to be complicated by

contradictions in the sources.
APZ (cap. 23) narrates that Osman was alive during the conquest of Bursa in 726
but didn’t participate because of gout and primarily because he wanted to see Orhan
win glory of his own. The first Ottoman coin, struck by Orhan in Bursa in 727 pro-
vides a terminus ante quem. A Byzantine short chronicle gives an exact date for the
fall of Bursa: 6 April 1326. The Anonymous (p. 20f.) adds that Osman died in 727
after reigning for nineteen years.
So far, so good, but the source closest to the events, Ibn Battuta who visited Bursa
in 1332, muddies the waters as he writes about Orhan: (p. 452): “It was his father
who captured the city of Bursa from the hands of the Greeks and his tomb is in the
mosque which was formerly a church of the Christians. It is told that he besieged the
city of Yaznik [Nicaea] for about twenty years but died before it was taken. Then this
son of his besieged it for twelve years before capturing it.” By this account Osman
himself took Bursa and besieged Yaznik for about twenty years which can be calcu-
lated: Iznik fell in 1331, so Orhan had been attacking it since 1319 and Osman before
him since about 1299, a momentous date that marks the beginning of Ottoman
independence. This would have Osman dying around 1319, incompatible with his
taking Bursa. Plainly, the traveler has confused Bursa and Iznik in a way that is diffi-
cult to sort out. The dates given by the Anonymous would have Osman coming to
power in 1308, a date which finds no support.
A document comes to the rescue: the endowment deed of the hospice at Mekece
is dated March 1324 (Beldiceanu 1967, 88; Lowry 2003, 75–8 with text). It is
headed by the tughra of Orhan who plainly is in charge, indicating that Osman
was no longer ruling and presumably deceased by that time. Mme Beldiceanu (p.
372) narrows the options by adducing the figure of Osman’s wife Malhatun who is
supposed to have died three months before Osman (APZ 28). Since she appears in
the Mekece document, Osman would have died between March 1324 and April
1326. One of the witnesses at Mekece was in fact Mal khatun daughter of Umar
beg and certainly a member of the Ottoman family like the other signatories who
include four brothers of Orhan and Osman’s daughter and granddaughter. But was
this Osman’s wife? Not certain, for the tradition names that wife as Malhun daugh-
ter of Sheikh Edebali (APZ 4, 28).
Since it is impossible iron out all the contradictions, it is probably safest to pre-
sume that Osman died a few years before or after 1324, so best to use “c.1320” or
“c.1324,” dates that also mark the accession of Orhan.
Notes
2. See p. 113.
3. See the discussions of Menteşe and Kastamonu, next.
5. Note that Gregoras omits Kastamonu and Menteşe, and presents a major
anachronism by including Sasa who disappeared from the scene in 1307 and
Saruhan who didn’t capture his capital Magnesia until 1313.
6. For relations between the Christian powers and the emirates of Menteşe and
Aydın, with analysis of the trade and texts of treaties, see Zachariadou 1983.
7. Osman’s lands and career are discussed in greater topographical detail in Chapter 1.
8. On this and subsequent dates, see p. 920, n. 47.
9. Neşri 71–3; Haci Bektaş: Taeschner 1928, 101 n. 1.
10. Neşri 65–9 reports the conquest of Karacahisar by Ertuğrul together with
Sultan Ala ad-Din and its loss after the death of the sultan (1237); this seems a
fiction to justify Osman’s much later occupation of the place (APZ cap. 6); the
resemblance of the two accounts of the conquest suggests that the earlier is just
a retrojection of the later.
11. For the geography of Eskişehir and its history in this period, see Lindner
2007, 57–67.
12. Al-Umari 350 from Balban. Its location between the lands of Suleyman Pasha
(Paphlagonia) and Germiyan leaves no doubt that Sultanönü/Eskişehir is meant.
13. For the date, see Wittek 1932, 351–4 but note that according to Cantacuzene
III.284, the Ottomans conquered Ankara in 1364, together with Gerede.
15. Beldiceanu 2000b; see above, n. 11.
16. For Karacahisar and the problems it poses, with full consideration of the
sources, see Lindner 2007, 67–80, and for its history in the Seljuk and Ottoman
periods, Doğru 2001.
17. Beldiceanu-Steinherr 1967, 86 n. 8. See also Chapter 4.

18. Toğan 1970, 331; cf. Turan 1971, 623–5. See Rashid al-Din 3, 643: Sulemish
granted the title of amir and banners to his followers who were drawn from
Syria and the frontiers.
19. Iznik in the title, Iznikmid in the text of the chapter.
20. al-Umari 340.
22. For a comprehensive chronology, see Danişmend 1947, 1–26; he often diverges
considerably from the traditional dates.
23. But note that Mme Beldiceanu (2003, 367 n. 119) would place the conquest of
Bilecik and Yarhisar in 1309 or 1310.
24. Attested by both Pachymeres XI.21 (4.455) and APZ cap. 20 in a rare coincidence
of sources. The document is discussed in more detail in Chapter 4.
25. For the full history of Germiyan, with relevant texts, see Varlık 1974.
26. RCEA 13, 5080.
27. Destan 21f.
28. Aflaki 661–3. The passage about Birgi seems to indicate that it was then under
Yakub of Germiyan’s control (through his subashi).
29. Ibid.
30. Pachymeres XI.21 (4.463); XI.23 (4.469); XI.25 (4.479).
31. Gregoras I.214.
32. A point made by Mme Beldiceanu 1984, 20f.; cf. Flemming 1964, 29f.
33. See below, p. 181.
34. The date is discussed (see Beldiceanu 1984, 20f.) but the main source, Aksarayi
251f. makes it clear that Choban’s expedition was to bring the Turcomans to
order, following the death of Uljaytu in 1316.
35. Uşak: Varlık 1974, 43f., 143f.; Sandıklı: RCEA 14, 5517.
36. See Chapter 7, pp. 208–211.
37. Yücel 1980, 33–51; cf. Cahen 1971.
38. Marino Sanudo ap. Hopf 144f.; for Licario, see Pachymeres V.27 (2.525–9). The
Turk’s surname, Caraman, might suggest that he had a connection with Germiyan.
39. Aksarayi 251f,; for the date, see note 00.
40. See Zachariadou 1997.
41. See 000.
42. Yücel 1980, 53–60.
43. Diler 2006, 359, 399; Ender 2003, 33–9.
44. See Cahen 1974, chapter XI, 44, 48.
45. Korobeinikov 2014, 274–80 and in its full complexity idem 2004; these events
are all very obscure.
46. Pachymeres II.6 (1.139f.). For the history of Philadelphia 1290–1390, see
Ahrweiler 1983 which draws on an impressive array of sources.

48. Beldiceanu-Steinherr 1984, 17–22; Ahrweiler 1983, 184, 190; see Chapter 2,
n. 00.
49. Gregoras I.360–362; Schreiner 1969, 389–95. For the local production,
Beldiceanu-Steinherr 1984, 29–34, Schreiner 1969, 411f.
50. See Ahrweiler 1983, 184.
51. He was Manuel Tagaris, a well-known imperial official.
53. Kalames apparently died before the emirate was established, for it bears his
son’s name.
54. For the history of the emirate, see Uzunçarşılı 1969, 96–103 and Zachariadou
1993b.
55. Pachymeres XI.26 (4.481), XI.27 (4.487f.) In both cases—as in the examples
that follow—the author refers only to “Persians” (i.e., Turks) without reference
to specific tribes or leaders.
56. Pachymeres X.26 (4.369), XI.21, 23 (4.465–9).
57. Pachymeres XII.34 (4.609).
58. Ender 2000, 21f.
59. But Saruhan Bey (c.1310–1345) did issue Western-style silver coins in the
1330s which will be discussed in Chapter 6.
60. See, below, p. 196.
61. See the inscriptions published by Wittek in Riefstahl 102–6.
63. Ender 2000, 173f., who lists these under Menteşe though they must be of Aydın.
64. See Lemerle 1957, 40–50 for Smyrna in this period.
65. Foss 1979, 145.
66. Korobeinikov 2014, 253.
67. Marino Sanudo in Hopf, 1873, 145; Pachymeres VI.20–21 (2.591–9).
68. Korobeinikov 2014, 263.
69. Uzunçarşılı 1969, 70–83.
70. See Wittek 1934, 24–53 for the obscure early history of Menteşe; cf.
Korobeinikov 2014, 254–6 with a speculative identification of Menteşe as a
Turkish-Kurdish tribal confederation from eastern Anatolia.
71. Pachymeres VII.26 (3.81f.).
72. For the abandonment of the Byzantine fleet and its consequences, see Ahrweiler
1966, 374–81.
73. Wittek 1934, 46f., 58.
74. Pachymeres XIII.13 (4.647f). Sasan is discussed in more detail in Chapter 4.
75. Gregoras I.214.
76. Wittek 1934, 54 prefers to identify this figure as the son of Menteşe, mentioned
in a document of 1366.
77. Zachariadou 11f.

78. Wittek 65, cf. Zachariadou 1983, 14.
79. See above, p. 180 and Foss 2019 p. 193.
80. See Cahen 1974, 42f.
82. Like the previous chapter, “The Overlords,” this section is a compilation from
standard works, intended to provide background information.
83. The course of Byzantine-Ilkhanid relations during the previous twenty years is
unknown: Korobeinikov 2014, 210.
84. Aksarayi 251f.
7
Western Asia Minor in the 1330s
By a remarkable coincidence, three sources that describe the emirates in

some detail are all contemporary, dating from the reign of Orhan:
Ibn Battuta (1304–1368 or 1377), was a native of Morocco to which he
always returned from travels that took him to most of the known world. His
family was traditionally associated with the office of kadi which he held
himself toward end of his life. He is known only from the travels which
brought him fame, high office, and adventure: these included the inner
Islamic lands plus Central Asia, India, China, Constantinople, and Black
Africa. In 1332 he visited Asia Minor where his account of the various
Turcoman states provides a unique level of personal information.1
Shihab al-Din Ahmad ibn Fadl Allah al-Umari (1301–1349), born in
Damascus, became a high-level administrator who rose to be head of chan-
cery in Cairo and Damascus, but was dismissed from both and imprisoned.
He was a prolific writer, noted for his detailed analysis of the organization of
the Mamluk empire and for his encyclopedic Masālik al-abṣār fī mamālik
al-amṣār which incorporates the works of Haydar al-Uryan of Sivrihisar in
Anatolia, who returned from there in 1333, and Balban the Genoese (born
Domenico Doria) whose information also dates from the 1330s. Umari
gives the date of al-Uryan’s visit and notes that he met Balban while they
were both in prison, so between 1337 and 1339. It is important to distin-
guish the summary account of al-Uryan from the more detailed description
of Balban, rather than to subsume them both under al-Umari.2 Neither
account, though, is as personal or vivid as that of ibn Battuta.
A unique source, an epic poem the Dusturname, presents the story of
one ruler, Umur of Aydın (1326–1348) who gained renown for his spectac-
ular campaigns against the Christian lands of Greece and Thrace. Little is
known of its author, Enveri, who participated in Mehmet II’s expeditions in
the Balkans and finished his epic in 1465, drawing on lost earlier sources.
He covered the period from 1317 to 1348 in some detail, focusing on Aydın
but bringing in the neighboring emirates as necessary.3
European sources provide important political and economic informa-
tion, notable among them the six treaties concluded between the Venetian
DOI: 10.1093/oso/9780198865438.003.0008
governor of Crete (and, on occasion, other European powers) with the rulers
of Menteşe and Aydın. They date from 1331 to 1353.4
Francesco Pegolotti (c.1290–1347), Florentine politician and merchant,
worked in Antwerp, London, and Cyprus. He produced a manual, La prat-
ica della mercatura between 1335 and 1343 which gives practical informa-
tion about all the places reached by Italian trade, as far away as Iran, Russia,
and China but predominantly the Mediterranean.5
Coins and inscriptions will be adduced to complete the range of contem-
porary sources.6
The following conspectus will present a quick view of the emirates, their
cities and military strength as presented by the three contemporaries.7
Deformed place names—common in al-Umari—have been corrected wher-
ever possible; those in quotation marks could not be resolved (Table 7.1).
The three sources agree to a considerable degree about the number of
emirates and their major cities, but the estimates of their armed forces vary
so widely that it seems possible only to treat them as representing orders of
Table 7.1
Ibn Battuta Al-Uryan Balban
Menteşe Milas 3,000 Mughla 100,000 (!)

Mughla, Milas
Aydın Berki 10,000 Berki 70,000
Birgi, Ayasuluk, Yazmir
Saruhan “Kasberdik” 8,000 Nif 8,000
Maghnisiya Maghnisiya 10,000
Karesi Balikesir, Bergama Akira, Bergama
Barghama, Bali Kasri 200,000 (!) 20,000
— — Sultan Önü8
Ottoman (Orhan) 25,000 Bursa 40,000
Bursa, Yaznik, Makaja, Kawiya,
Yanija, Kainuk,
Muturni, Buli
— Kainuk Hisar 3,000 —
Gerede Kerdeleh 3,000 —
Garadaibuli
“Qawiya”
“Qawiya” 14,000
Candaroğlu Kastamonu, Sinop, Kastamonu, Sinop
Burlu, Qastamuniya, Sanub “Bura” 25,000
30,000
Western Asia Minor in the 1330s 193
magnitude, with some very large and some very small states. Note that Ibn
Battuta calls all the rulers “sultan” and their governors “amir.” For al-Umari’s
sources each beylik is a “memleket”; both identify the states by the name of
their capitals.
Menteşe
Ibn Battuta 428 first stopped in Mughla where he met Ibrahim Beg, son of
the sultan of Milas, then (428–430) proceeded to Milas which he considered
one of the finest, largest cities of Rum; its sultan Shuja al-Din Urkhan Beg
son of al-Mantasha, one of the best of princes. A body of Doctor of Law was
always at his court. Menteşe at this time was on bad terms with Ayasuluk.
The Sultan’s residence was in Barjin (Peçin) two miles from Milas, a new
place with fine buildings and mosques (Fig. 7.1). When Ibn Battuta passed
through, the congregational mosque had not yet been finished, but an
inscription shows that it was completed in 1332.9
In Milas, as in most of the places where he stayed in Asia Minor, Ibn
Battuta enjoyed the hospitality of the Young Akhis, members of an organi-
zation of merchants and artisans who offered hospitality to strangers,
defended the local population against tyrants and criminals, and even acted
as governors in cities where there was no sultan.10
Al-Uryan 21/339 sparsely notes that the sons of Menteşe rule Milas and
have only 3,000 horsemen but adds the detail that the emir of Finike (in
Lycia) who rules in the name of the sovereign of Antalya (Haydar, son of
Fig. 7.1 Peçin, Menteşe’s new capital

Dundar of the Hamidoğulları) himself belongs to the family of Menteşe.

This may reflect conquest of some of Menteşe’s lands by Hamid in the trou-
bles that followed the death of Uljaytu in 1316, or a marriage alliance
between the two neighboring emirates which often cooperated in revolts
against the Mongols.11
Balban 47f./370: The memleket of Mughla was ruled by Orkhan son of
Menteşe.12 It had 50 cities, 200 forts. and an (implausibly large) army of
100,000. Its army constantly fights infidels and Muslims (something the
sources—especially the Ottoman—do not usually admit). On ship or horse;
he is in arms night and day, never rests, and is victorious virtually every-
where. Among Turk emirs, only Germiyan tries to gain his friendship and
good will; Germiyan is the only prince whose superiority Menteşe recog-
nizes; all the rest are considered inferior by rank.
Western
Two treaties of this period between the Venetian governor of Crete and
Menteşe have survived. The treaty of 1331, concluded with Orhan, provided
protection for merchants of both sides and land to the Italians for a church
and houses in Palatia, Menteşe’s capital, where a Venetian consul was
already established.13 Here, the Italians bought horses, cattle, sheep, leather,
carpets, grain and sold wine, soap, and textiles.14 This treaty was followed
by another in 1337, concluded with Ibrahim, which dealt in more detail
with freedom of navigation, solution of disputes and maintenance of
peace—all to apply to a large region of Greece controlled by Venice. The
Duke of Crete promised not to collaborate or associate with the enemies of
Menteşe. An array of subordinate rulers also swore to obey the treaty:
Ibrahim’s brother Hızır of Çine, Elyas Beg of Tavas, Melik-Eshref of Makri,
and the kadis of Palatia and Milas.15 The treaty was renewed in 1353.16
Inclusion of Elyas Beg of Tavas may indicate a recent expansion of
Menteşe, for when Ibn Battuta visited the castle, heavily fortified against
robbers (presumably Turcomans), it was under the command of Ilyas Beg,
who is not mentioned as subordinate to Menteşe.17 Balban described “Tawaza”
as being under an independent prince who had four fortresses and 600
towns, with an army of 4,000 cavalry and 10,000 foot.18 Control of Tavas
would mean that Menteşe’s power extended far inland, to the lands domi-
nated by Turcoman nomads.
Pegolotti mentions Palatia only briefly, as a source of grain and as one of

the ports from which the alum of Kütahya was shipped.19
Inscriptions
Inscriptions attest to the construction of a mosque in Milas and of a grand

congregational mosque in the new capital Peçin.20
Coins
Orkhan (c.1320–1335) struck coins in Milas—in a normal beylik type.21

His mint in Palatia produced remarkable silver coins of a purely western
type. They portray an enthroned king holding a scepter and globus cross on
the obverse and a cross entwined with lilies on the reverse. Struck in fine
silver to a standard of around 3.75 grams and averaging 29 mm in diameter,
they correspond in every respect to the so-called gigliati of Robert of Anjou,
king of Naples (1309–1343), but their inscriptions make it clear that they
were issues of Orkhan of Menteşe:
HANC MONETAM F:IER[I] I[VSSI]T VRCN: I: E

MANDAVIT DOMINVS PA:LATIE
These were evidently issued to facilitate the extensive trade between

Menteşe and the Christian states of the Aegean and Mediterranean, but they
seem to have been an experiment that failed, for they were struck in exceed-
ingly small quantities. Much more common were direct imitations of the
Neapolitan coins, struck perhaps in Palatia.22 Aydın and Saruhan produced
similar gigliati.
Menteşe was a rich and successful state whose extensive trade was sup-
plemented by raiding. Successful raiding brought revenue from Greek
states: in 1332 Menteşe imposed tribute on the duke of Naxos and by 1337
also on Negroponte (Euboea). However, profitable joint expeditions with
Aydın came to an end in 1332 as the emirates began to quarrel.23
Its wealth enabled Menteşe to maintain an army and a fleet and to build
an entirely new capital city. Its sultan ruled from Milas, his son’s headquar-
ters was in Mughla while other relatives commanded three other capitals.
Palatia, the ancient Miletus, the economic center of the beylik and the place
where gigliati were struck was evidently under the sultan’s direct control. In
terms of wealth, Menteşe was apparently second only to Aydın and in prestige
inferior only to Germiyan.
Aydın
When ibn Battuta (438–447) came to Aydin’s capital Birgi (Fig. 7.2), Mehmet
son of Aydın (1307–1334) entertained him lavishly, first in the mountain
resort (the modern Bozdağ) where he spent the summer, enjoying the
cool streams and the shade of the walnut trees (Fig. 7.3). The sultan,
whom ibn Battuta described as one of the best, most generous, and wor-
thiest, sent his sons Hızır and Umur to greet the traveler and provide him
with tents and rugs, asking him to write down hadiths that he knew, for
the sultan was an avid patron of Koranic learning. After a long stay in the
mountain, ibn Battuta and Mehmet descended to his palace in Birgi to be
greeted by elegantly attired Greek pages. A long flight of stairs led up to
the audience hall that had a pool with bronze lions at each corner spout-
ing water and raised benches covered with carpets. Koran readers were in
constant attendance. After lavish banquets, where food was served in gold
Fig. 7.2 Birgi, Aydın’s capital [50]

Fig. 7.3 Aydın’s summer resort
and silver bowls, and loaded with gifts that included a hundred gold
dinars and a thousand silver dirhams, Ibn Battuta departed, a Greek slave
added to his retinue by the sultan. He described the other centers of the
emirate, ruled by Mehmet’s sons:
Tire: a fine town with running streams, gardens, and fruits. Ibn Battuta
did not meet the governor, Mehmet’s younger son Suleiman, who had fled
to his father-in-law Orhan of Menteşe.
Ayasuluk: Governor Hidir Beg: Its congregational mosque, converted
from a Greek church was one of the most beautiful mosques in the world,
with its walls and floor of marble and pools under its eleven domes. The city
had fifteen gates and was traversed by a river bordered by trees, vines, and
trellises of jasmine (Fig. 7.4).
Yazmir: mostly in ruins (probably from the 1327–1329 war when Umur
captured the harbor fortress from the Christians). Its amir the generous
pious Omar Beg was constantly involved in jihad against the Christians
whom he attacked with his galleys, ravaging the lands of Constantinople.
After a successful expedition, he would give away all the loot, then have to
fight another campaign.
According to al-Uryan 21/339, Mehmet son of Aydın who ruled from
Birgi, had some 10,000 horses but lived in compete isolation without friends
or allies; a mysterious description.24
Fig. 7.4 The citadel of Ephesus
Balban 45f./369: reported that the state was ruled by sons of Aydın; its
capital was Berki. It had 60 cities, 300 forts, and an army of 70,000 horse-
men, all warlike, skilled with the sword and lance. They have fought memo-
rable wars against Greeks, Franks, and other infidels.
Western Sources
Trade was as important as raiding and doubtless a more dependable source

of wealth (Fig. 7.5). Its main entrepôt was Ayasuluk, the ancient Ephesus,
which for Pegolotti was one of the major ports of Turchia.25 Part of its huge
church had been turned into a market, where silk, wool, wheat, and other
products were on sale. Slaves were also an important item of commerce.
Altoluogo, as the Italians called the city, provided a market for gaudy
European fabrics: azure, vermilion, emerald, pistachio, and turquoise cloth
were especial in demand, as were Florentine dyes for wool. In return, the
Turks sold alum, grain, wax, rice, and hemp.26 The alum was produced near
Kütahya in Germiyan and exported through Altoluogo. Hızır’s capital
traded with Pisa, Florence, Venice, Constantinople, Cyprus, and Rhodes.27
This rich trade naturally needed to be regulated, so in 1337, for the first
time, Aydın and the Venetians of Crete made a treaty: it was concluded with
Fig. 7.5 The Cayster valley, one source of Aydın’s wealth
“Çelebi” lord and emir of Theologos, his brother Umur and his other
brothers, the sons of Aydın. It provided for an Italian consul at Theologos,
land in the city for Italian merchants and a promise that Aydın would not
put ships to sea during the duration of the treaty.28
Coins
Only a few of Aydın’s rulers issued coins. Mehmet (1307–1334), Ibrahim

(1334–1356?), and Suleyman (1334–1349)—all of whom reigned in the
Cayster valley—issued silver pieces without mint or date; all are extremely
rare. Western style gigliati struck in Ayasuluk (Theologos) do not bear the
sovereign’s name but were probably issued by order of Hızır (1334–1360),
closely involved in trade with the West (Fig. 7.6).
They are inscribed:
MONETA.QVE.FIT.IN.ThEOLOGOS/DE MANDATO DNI EIVSDE[M]
LOCI.
A unique gold piece, modeled on that of Florence, was a product of the
same mint.
Fig. 7.6 Silver gigliato of Theologos (Ephesus). Courtesy of the Princeton

University Numismatic Collection, Department of Rare Books and Special
Collections, Firestone Library
Inscriptions
Inscriptions attest the building of Mehmet’s tomb in Birgi in 1334 and two
mosques in Tire in 1338.29
Umur Bey
Umur of Izmir, alone among all these beys, was the subject of an epic poem,
known as the Destan, which in 2514 verses describes his life and exploits in
heroic terms, with full details of his attacks on the infidels of Greece, Thrace,
and the islands.
In 1326 or 1327, Mehmet Ibn Aydın divided his lands among his sons, an
activity characteristic of other emirates.30 Hızır, the eldest, received Ayasuluk
and Sultan Hisar, lands at both ends of the emirate, evidently to protect the
frontiers. Assigning lands so widely separated seems anomalous, but note
the coins struck in the name of Uljaytu in 1310 in precisely these two
places.31 In fact, there is a precedent for such an assignment: the Mongols
normally gave their eldest sons the lands farthest from the capital, presum-
ably to defend or advance the frontiers.32 The Ilkhans in particular sent the
heir to the throne to govern the unruly and remote frontier district of
Khorasan.33 Mongol influence might not have been out of place here, for
Aydın was (or had been) subordinate to Germiyan which in its turn had
submitted to the Ilkhans.
In any case, the second son Umur got Izmir, Ibrahim Bodemya,34 and
Suleyman Tire, while Mehmet himself stayed in Birgi. The major Byzantine
center Tralles, now called Aydın, does not appear. It apparently lay in ruins
since its capture by Menteşe in 1284.35 This division did not imply the
breakup of the emirate, for the brothers continued to cooperate, even after
the death of their father in 1334. Sultan Mehmet, despite his remote loca-
tion on the slopes of the mountain, far from the scenes of action, had
enjoyed great success and wealth. He had already built the first monumental
mosque of any of these emirates. Not entirely in vain did he claim the gran-
diose titles that adorned his tomb there: ‘the great sultan, the fighter of holy
war, defender of the faith, founder of pious works, sultan of the ghazis,
fighter for the state and the religion.”36
The two decades from his conquest of Izmir in 1329 brought Umur to the
height of his power and renown.37 In 1332, he attacked Gallipoli, the islands,
and the Peloponnesus. He renewed his operations in the Peloponnesus in
1335. In the same year, he led an attack on long-suffering Philadelphia,
described as the point of passage between Germiyan, Saruhan and Aydın.
Umur didn’t take the city, but extracted tribute and withdrew. Soon after, he
met with Cantacuzene and agreed to renounce the tribute. Now cooperating
with Byzantium, he sent a mercenary force into Albania in 1337, moved
against crusader-occupied Greece two years later and in 1341 even sent a fleet
to the mouth of the Danube. All this required money, ships, and manpower.
After 1329, when he gained full control of Izmir, Umur made it his arsenal
for constructing fleets and his base for launching them against Thrace and
Greece.38 The numbers were impressive, rising from 28 ships in 1330 to 250
two years later, 270 in 1335 and 350 in 1341. Some of these totals at least
would have included ships from Saruhan and Karesi which joined in the
expeditions.39 Similarly, the manpower involved included warriors eager not
only for plunder but to participate as ghazis against the infidel. By 1343, Umur
could command a force of 15,000. These large numbers in turn provoked a
need for campaigns that would bring back loot or involve lucrative service as
mercenaries, especially when he was in alliance with Cantacuzene.40
By all accounts, Aydın was the richest and most active of these states, its
wealth from trade centered on its great port of Ayasuluk and from the
extensive raiding of the most glorious figure of the age, Umur of Izmir. A
contemporary observer, overwhelmed by the manifestations of wealth
throughout the emirate could reasonably have supposed that Aydın was
destined for a great future as the leading Turcoman state. Four brothers
ruled from four capitals, all in cooperation and recognizing Umur as the
leader. But note that Umur spent money as soon as he acquired it and that
his expeditions were aimed at loot, not conquest. Like most of these emir-
ates the frontiers of Aydın appear to have remained stable, with no territory
gained or lost.
Philadelphia
The anomalous Byzantine city of Philadelphia prospered through this

decade, despite the inevitable problems of attack by its neighbors and con-
sequent tribute to be paid to them. Its relation to the Empire, from which it
was cut off by the adjacent emirates, has not been determined.
In any case, it attracted the attention of Umur of Aydin who in 1335
attacked Philadelphia. The Destan reports the event in terms that seem to
claim that, after a long siege, Umur captured and looted the city and installed
a garrison. Its tekfur supposedly surrendered and a messenger rushed off to
give the bad news to the tekfur of Istanbul.41 Alternatively, the text could
mean that Philadelphia surrendered on terms, paying a huge ransom. This
seems the more likely, for not long after Cantacuzene came to Clazomenae
west of Izmir and made a deal by which Umur agreed to cancel the tribute
and to become an ally of the Empire. The terms were confirmed by the
emperor in person, meeting Umur off the mountainous peninsula of Kara
Burun west of Izmir.42
Pegolotti, writing around this time, mentions the “Perperi di Filadelfia,”
i.e., gold coins struck in Philadelphia. These coins have not been identified;
if the information is correct, it would indicate a high level of prosperity.43
Saruhan
Ibn Battuta (447f.) called Magnesia a large fine city on a hill with a rich
fertile plain. Fuja (Phocaea), a strongly fortified infidel city on the coast a
day’s journey away sent gifts every year to the sultan, who in turn left
them alone.
On their way across the mountains from Magnisia to Bergama, Ibn
Battuta’s party encountered a Turcoman camp and had to stand watch all
night against robbery.
Al-Uryan 21/339 calls this beylik “Kasberdik,” a name that has not been
resolved. It had 8,000 horsemen.
Balban 44f./367–9) divides his discussion into two chapters, Nif (the
ancient Nymphaeum) and Manisa:
Nif ’s ruler Ali pasha brother of Saruhan had eight cities and 30 forts.
His 8,000 horses were joined by a flood of infantry who fought with
bows and javelins. The province extended along the mountain heights
(Figs. 7.7, 7.8).
Saruhan ruling from Manisa had 15 cities and 20 forts with an army of
10,000 horses, all warlike. The inhabitants mounted frequent naval expedi-
tions, never letting themselves be surprised. Both states were distinguished
by their wars against the infidels.
The only coinage attributable to this period, issued by Saruhan Bey
(1300–1345) consists of two types of gigliati like those of Menteşe and Aydın,
inscribed
MONETA: QVE: FIT: MNGLASIE

DE: VOLVNTE: DNI: EISDEM: [L]O[C]I
or: M[ONETA MANGLA]SIE SARCANI
DE VOL[VNTAT]E DNI EISD LOCI
They are extremely rare.
Fig. 7.7 Nif/Nymphaeum: Byzantine and later fortifications [55]

Fig. 7.8 Nymphaeum: the Lascarid palace
The rich emirate of Saruhan receives only scanty notice.44 It came to the
attention of Byzantium when it was attacking imperial territory, often in
alliance with one of its neighbors. Ibn Battuta and al-Umari provide the
only information about the emirate itself.
Byzantine sources reflect the changing relations between emirate and
empire. It was probably in 1332 that Suleyman, son of Saruhan organized an
attack on Gallipoli together with Umur and the son of the emir of Menteşe;
they returned after capturing a small coastal fort.45 In 1335, the emperor
Andronicus III came in person to Phocaea to seek help against the Genoese
who had taken Mitylene. From there, he sent a message to Saruhan “who
ruled the lands east of Phocaea.” The emir met the emperor and concluded
an alliance, promising men and ships to attack Mitylene and the strategic
Genoese New Phocaea. Saruhan’s twenty-four ships, together with forces
provided by Umur, prevailed: the Genoese returned Mitylene but were
allowed to keep New Phocaea under Byzantine sovereignty. Saruhan in turn
promised to continue to supply New Phocaea with foodstuffs.46
In 1341, “Sarkhanes satrap of Lydia” joined Yahşi of Pergamum in attacking
Thrace where they were severely defeated. In the Byzantine riposte, Saruhan’s
coastland was ravaged and one (unnamed) town captured, producing many
slaves.47
Saruhan was evidently a middling power, oriented toward the west

through trade (as attested by the gigliati, one of which was found in the
Sardis excavations) or raiding, often in alliance with the other maritime
states. Saruhan had an army and a fleet, both of moderate size. Its efforts
were especially directed toward the Genoese base of New Phocaea which
sometimes brought it to cooperate with Byzantium. It was a family affair,
with the main ruler in Manisa and his brother in Nymphaeum. The sources
say nothing about Saruhan’s control of the Hermus valley, one of the richest
in Asia Minor, but their writ hardly extended to the mountains on the north
where there was danger from nomad Turcomans.
Karesi
Ibn Battuta (448f.) described Bergama as a city in ruins with a great fortress
(Fig. 7.9); he didn’t meet Yahşi Khan who was in his summer campgrounds.
He did meet Dumur Khan (whom he considered worthless) and visited the
fine populous city of Balikesir, built by the sultan’s father. Its congregational
mosque was still lacking a roof. Ibn Battuta notes that these rulers used the
title khan, more exalted than the usual beg of the other emirs.48
Fig. 7.9 Bergama, a Karesi capital

Fig. 7.10 Achyraous/“Akira,” Karesi’s first capital
For al-Uryan (22/339), the principality of Balıkesir belonged to Demirhan

son of Karesi, with a subordinate (whose name, “Senbogha” has been gar-
bled) governing Pergamum, two days away. The whole country was fertile,
easy to defend, and comprised a vast territory. The army was very small,
with no more than 200 horsemen, and yet the ruler had no fear, relying on
the strong position of his country, which no one would dare to try to
conquer.
Balban (43f./365f.) describes two states: “Akira” (Achyraous), ruled by
Demirkhan son of Karesi, and “Marmara”, the domain of his brother Yahşi49
(Fig. 7.10). Demirkhan was a neighbor of Orhan; his domain crossed by
travelers on their way to Sinope. Its cities, fortresses and troops were more
numerous than those of Orhan; its inhabitants had more strength and
energy. Its warlike ruler, who never let himself be surprised, constantly
fought with the Greeks, putting many ships to sea. The state produced enor-
mous quantities of silk and laudanum which they exported to the Greeks.
Yahşi of “Marmara,” son of Karesi and brother of Demirkhan, ruled 15
cities and 15 forts, all on high mountains next to the sea. He had 20,000
horses and was constantly at war with the Greeks, equipping fleets to ravage
their lands. The brave warlike inhabitants fought without respite, not
stopped by city walls or deserts. They captured many young Greeks and
daughters of the “Khozars.” Large numbers of slaves were captured in war;
slave dealers arrived every day.
Western
Pegolotti (24) reported that: “Lupai” (i.e., Lopadium) produced alum of

middling quality which was exported through the port of Triglia on the
Marmora, a town also noted for its wine.50
The Karesi family appears in history in 1328 when Andronicus III came to
Cyzicus and Pegae to negotiate with the “ruler of Phrygia,” Temirhan son of
Yahşi, about the imperial possessions in the Hellespont region, which Karesi
had been harassing.51 The discussions, during which Temirhan symbolically
submitted to the emperor, produced a promise by the emir not to attack
Byzantine territories—a promise that was kept. Temirhan (1327–1335) is
usually considered to have been the brother, not the son, of Yahşi
(1327–1343). It appears, then, that the senior branch of the family, repre-
sented by Temirhan, ruled interior Mysia from Balikesir, while Yahşi and
his family, with their capital in Pergamum, controlled the coastlands, hence
their association with Troy, the Hellespont and the Marmora. Yahşi’s son
Suleyman appears to have been based in the Troad.52 Balıkesir itself was a
new capital, built up by Karesi bey.
Yahşi did indeed build fleets and used them to raid Thrace. By 1334 he
had taken control of Adramyttion and was planning, according to an Italian
source, to embark on piracy with a large fleet. It was crushed by a coalition
of the Christian powers.53 Nevertheless, in 1337, “the Turks of Troy and the
Hellespont” embarked their forces on ships and attacked Thrace.54
Coins
Each of the rulers of this dynasty struck coins, none of them bearing mint-
mark or date. Those of Suleyman, the son of Yahşi, uniquely feature a bust
on the obverse (otherwise coins of the emirates were aniconic) while those
of Yahşi give him—anomalously, for the written sources describe him as
subordinate to Demirhan—the grand title of amir al-a’zam and even al-
sultan al-a’zam.55 Note that the sequence of coins continues past 1343, into
the reign of the otherwise unknown Beylerbeyi Çelebi, who apparently suc-
ceeded Yahşi. His akçe is identical in type to one of the issues of Orhan,
suggesting a close monetary association on the eve of the Ottoman con-
quest.56 It is conceivable that he is the “Haci Ilbeyi” of APZ. Both have titles,
not real names, and “Haci Ilbeyi” is supposed to have gone on ruling
Bergama after the rest of the emirate had fallen to Orhan, just as Beylerbeyi
Çelebi seems to have been the last emir of Karesi.57
Although it ruled a large and strategic territory, with coasts of the sea on
Marmora and the Aegean, the history of Karesi is poorly known. It was
ruled by two brothers; the senior had a newly built capital of Balıkesir in the
interior while his colleague of Bergama controlled the coastlands. Both had
close relations with Byzantium, friendly or hostile, and gained the reputation
of successful fighters against the infidel Greeks. Although it appears to have
been a middling power, Karesi had a reputation of strength, easily able to
defend itself and more powerful than its neighbor, Orhan. As it turned out,
both al-Uryan and Balban were totally mistaken.
Germiyan
Ibn Battuta (424f) never visited Germiyan, for the Sultan of Göl Hisar sent a
body of horsemen to escort his party to Ladhiq because the plain was
infested by troops brigands called al-Jarmiyan. They were said to be descen-
dants of Yazid ibn Muawiya and they had a city called Kutahiya. “God
preserved us from them,” the traveler concluded.
For al-Uryan (22/340), Germiyan son of Alishir (? text has Ghadshahr)
ruled from Kütahya. His 40,000 horsemen, constantly trained in combat,
were practically invincible. He enjoyed absolute authority. The Turkish
emirs hated him and did all they could to destroy him.
According to Balban (30f./348f.), all the Turkish emirs recognized the
authority of the ruler of Germiyan rendering him in many ways the
honors due a sultan. Some pay a fixed tribute; others send gifts. They
turn to him when they have trouble and are happy to follow his advice;
they use his support to gain advantage over each other. They receive
from him robes, presents, appointments, and marks of honor. Although
he cannot appoint or dismiss them, he has unchallenged authority. But
his relations with them are like those that existed between the last
caliphs and the rulers of the various states. They are obliged to use all
the formulas of politesse in addressing him. He has the most land, sub-
jects, and soldiers of all.
(33/351): Germiyan was the most important of sixteen principalities, the
closest to the lands of the family of Genghis Khan.
(34–36/354): Germiyan exercises sovereignty over the Turkish chiefs,
expands his lands at their expense. His capital was Kütahya, a large city with
an important citadel (Figs. 7.11, 7.12). His land was rich, well populated and
cultivated, with enormous flocks. He claimed to have 700 cities or fortresses;
Fig. 7.11 Kütahya: walls
Fig. 7.12 Kütahya citadel [60]
and could put into the field 200,000 horse, foot, lancers, and archers, all well
supplied, well-armed with Damascus steel. He had innumerable flocks and
the fastest horses.
The emperor of Constantinople pays a tribute of 100,000 gold coin of
Constantinople plus magnificent presents.
The ruler of Germiyan constantly inspects his troops, prepares for war.
He has a grand court: emirs, vezirs, kadis, secretaries, courtiers, and pages;
he has treasures, stables, kitchens, and palaces—the pomp and luxury suit-
able for a sultan.
Among his dependencies is the city Gümüs şehir, a rich mine better than
the Mongol; also an alum mine which brings great wealth, and a city
Sivrikoy (?) which produces only rice.
The inhabitants don’t always make distinction between what is allowed
and what not; they shed blood with indifference; their swords are covered
with the blood of enemies, their arrows pour onto adversaries.
Western
Pegolotti 43, 193, 369f: Kütahya was a major source of good quality alum,
which was exported through Theologos, but also through Palatia.
Coins
Coins give little hint of the wealth and power of Germiyan: Yaqub I
(c.1300–1340) struck one type of akce, apparently late in his reign, while
Mehmet (1340–1361) produced five types. None are dated; only one names
a mint which may possibly be read Simav.58 All are very rare.
Apparently no inscriptions have survived from this period.
By all accounts—except those of the Ottomans and their admirer Ibn
Battuta—Germiyan was and had long been the most powerful of these
states. It had a strategic location, broad and rich lands, and derived consid-
erable wealth from its mines of silver and alum. Its large and well-trained
army allowed it to exercise a dominating influence over the other emirates
who, however reluctantly, recognized its superiority. Although it was land-
locked, Germiyan had relations with Byzantium and received a large tribute
(which the Byzantine sources don’t mention). Its power and location made
it a desirable ally for Christian powers harassed by the maritime emirates:
in 1332 Venice was seriously considering such an alliance. In 1341,
Cantacuzene made an agreement with “Aliseres the satrap of Kotyaion” by
which Byzantium and Germiyan would jointly attack Saruhan.59
The sources only hint at the source of Germiyan’s authority but Balban’s
cryptic remark that it was closest to the lands of the Mongols may suggest
that it had Mongol backing and may even have been an agent of Ilkhanid
rule. None of this made Germiyan popular: they were widely feared and
hated as expressed in the warning Ibn Battuta received from his host the
emir of Gölhisar (in the lands of Hamid) against the bandits called
“al-Jarmiyan.”
Ottoman
For Ibn Battuta (449–460) Bursa was an important city, with fine bazaars,
wide streets, and hot springs (Fig. 7.13). Its sultan “Urhkan Bak son of
Othman Jik” was the greatest king of the Turkmen, the richest in wealth,
land, and military forces. Orhan had nearly 100 forts which he frequently
traveled to inspect. He constantly fought with the infidel, keeping them
under siege. His father had captured Bursa from the Greeks and besieged
Yaznik for 20 years, dying before it was taken; Orhan besieged it for another
12 years.60
He traveled to Yaznik (with an intermediate overnight stop at Kurluh)
through fertile country. The city, surrounded by water and approached over
a viaduct, was in moldering condition, uninhabited except for a few men in
the sultan’s service under the command of his wife Bayalun Khatun. It was
surrounded by four walls, with orchards and fields within.
Fig. 7.13 Bursa, Orhan’s capital

His stages on the road east reveal the extent of Orhan’s land: Makaja, the
Sangarius ford, Kawiya, Yanija a large fine township, Kainuk a small town
inhabited by Greeks, with no trees or vines, only saffron, Muturni, Buli.
These were all under Ottoman control, showing that their domain stretched
far east of the Sangarius into Paphlagonia.
Al-Uryan (22/339f.) reported that the “principality of Orhan son of
Osman” maintained 25,000 horses. His lands bordered the strait of
Constantinople with whose ruler he was constantly at war. Since Orhan
usually won, he was considered the most dangerous enemy of the Greek
emperor who paid him a monthly tribute.61 On one occasion, his forces
crossed by sea and ravaged the Christian lands.
According to Balban (421f./364f.) the ruler “Orhan son of Taman,”
had his capital at Bursa. He had 50 cities and even more fortresses, with
40,000 horses and innumerable foot. But his troops were not very warlike,
more impressive in appearance than in reality. He was peaceful toward
his neighbors, very inclined to help his allies, but constantly at war with
his numerous enemies. His subjects were ill-intentioned toward him; his
neighbors lived in open hostility. The population was treacherous, full of
hate and evil thoughts. The land had 300 hot springs which cured a range
of diseases.
(For Orhan’s coins, inscriptions, and documents, see Chapter 4.)
Unlike the maritime emirates, the Ottomans were not yet a naval power.
Orhan in the 1330s had only 36 light ships suitable for raiding but not for
any major expeditions.62 Nevertheless, all the contemporaries agree that
Orhan was one of the great powers of the day. Ibn Battuta accurately per-
ceived him to be the “greatest king of the Turkmen” while the anti-Ottoman
Balban was not impressed, believing that Karesi was stronger than its
Ottoman neighbor. Orhan led the fight against Byzantium and in this
decade completed his conquest of Bithynia with the capture of Nicaea in
1331 and Nicomedia in 1337. Unlike the others, he conquered territory and
held it, constantly increasing his power base.
The documentary record shows that the Ottomans, beginning with
Orhan’s conquest of maritime Bithynia in the 1330s, exploited their new
territories to provide endowments for the buildings that established an
Ottoman presence in the cities by means of buildings and institutions typical
of Islam—mosque, medrese, imaret, covered market—that at the same time
advertised their wealth and power.63
Göynük Hisar
al-Uryan (22/340) reports this small state as belonging to “Amer-Djakou”

who had 3,000 horses.
When Ibn Battuta passed through, it was under Ottoman rule. The name
of its emir indicates a Mongol origin, but nothing further is known.64
Gerede
For Ibn Battuta (460f.): Garadai Buli was a large fine city, with spacious
streets, bazaars, and separate quarters for each community. Its ruler, Shah
Bak, was of the “middling class of sultans in this country.”
Al-Uryan (22/340): names the ruler of “Gerdeleh” as Shahin; he had
3,000 horses.
Sultanönü
Mentioned only by Balban (32/350) apropos of the conquering expedition

of Timurtash in 1326: located between the lands of Suleiman Pasha and
Germiyan, it had no walled cities but large towns and vast plains.65
These small states somehow survived among their more powerful neigh-
bors who were destined to swallow them up. That was evidently the case of
Göynük on the route that led east from Geyve and some sixty-five kilome-
ters from the Sangarius. Al-Uryan seems to have caught it at the last
moment of independence, for when Ibn Battuta passed through, it was in
Ottoman hands as were Mudurnu and Bolu, the next stages on the route
east. Nothing further is known.
Gerede, about one hundred and forty kilometers east of Göynük at an
important road junction, seems to have been sizable and prosperous and
the seat of an independent small emirate, recorded as paying tax to the
Mongols in the 1330s. It was the neighbor of the Candaroğulları who may
have taken it by the 1340s, according to the evidence of a unique coin struck
there in a year ending in –4, presumably 744 (1343).66 The Ottomans in turn
conquered it in 1354 from the “eastern Scythians,” that is, the Eretnids.67
Cantacuzene, the source for this event, describes Gerede as one of the most
prominent towns of the Scythians. In any case, it was still independent

and prosperous in the 1330s, though squeezed between Ottomans and
Candaroğullari.
As for Sultan Önü, an object of Osman’s attention, it lay between Ottoman
and Germiyan lands, perhaps surviving through the protection of the latter.
“Qawiya”
It is hard to know what to make of “Qawiya” attested only by Balban

(41/363f.), who names its ruler as Murad al-Din Hamza.
It lay west of Samsun, north of the land of Suleyman Pasha (Katamonu)
and had the Kasis mountain on its west. Most of the travelers and merchants
of Egypt and Syria passed through Murad’s territories on their way to the
Black Sea. Its capital was “Qawiya,” whose prince possessed 10 cities and 10
forts and had a force of 7,000 horses and 7,000 feet, but for a campaign he
could put together a much larger force. His enemies don’t dare attack him.
The population was peaceful, with great affection for its rulers.
Mention of Samsun and the Black Sea suggest a location far to the east,
but note that Balban defines Ottoman lands as east of Murad, west of Samsun
and Sinope, with Mt Kasis as its western limit. In other words, this Qawiya
was an immediate neighbor of the Ottomans. This might suggest a location
in western Paphlagonia, which already contained the small states of Göynük
and Gerede between Ottoman territories and those of Kastamonu. But this
Qawiya is something larger, with a respectable armed force.
“Qawiya” could also be read as Geyve, a well- known place on the
Sangarius, but this lay well within the Ottoman lands. Balban sometimes
lists as two separate states family regimes where brothers rule separate parts,
so might a province of Geyve be such a case? Not likely since no Murad al-
din Hamza is known in the Ottoman family.
On the other hand, a certain Murad beg ruled the city of “Bura” under
Ibrahim of Kastamonu, of whose family he was presumably a member. The
mystery remains.
Kastamonu
Ibn Battuta (461–468) began his visit at Burlu (Safranbolu), a small town
with a citadel on a steep hill; its amir was Ali Bak son of Suleyman padshah
king of Qastamuniya which is one of the largest and finest cities (Fig. 7.14).
Fig. 7.14 Kastamonu: city center, with citadel in background
He praised the hospice at Taşköprü. The village had been dedicated for its
endowment while the bazaar was the endowment for the congregational
mosque, with generous provision for pilgrims.
For him, Sanub was a superb city surrounded by the sea, with only one
gate for which permission was required from the governor Ibrahim Bey son
of Suleyman Bey. On the hill were eleven villages of Greeks under the
Muslims, a hermitage, and a hospice with food for travelers. Ibn Battuta
described the city’s beautiful congregational mosque built by the Pervane,
who was succeeded by his son Gazi Çelebi, and then by sultan Suleyman.
Gazi Çelebi was famed as a frogman and for capturing an entire enemy fleet.
The locals, especially army officers, made much use of hashish.
Al-Uryan 23/340f.: Kastamonu was ruled by Ibrahim son of Suleyman; it
had an army of 30,000 horses or even more, and many fortresses and cities,
the most famous Sinope where an amir Gazi Çelebi rules for the prince;
also, the city of “Bura” governed by Murat beg. The prince had friendly rela-
tions with the rulers of Egypt and maintained correspondence with them.
The country is famous for its excellent horses, which fetch enormous prices
among the Arabs.
Balban 39f./361f.: It was ruled by Ibrahim son of Suleyman, who had
governed Sinope on the shore of the Black Sea, frequented by ships going to
Kipchak, Khazars, Rus, and Bulgars. Its capital is Kastamonu. Its sovereign
has forty important cities and fortresses and 25,000 horses of excellent
quality. He maintains friendly relations with the sultan of Egypt, constantly

exchanging ambassadors and gifts; he receives much help against his ene-
mies. At present he is under the authority of Melik Nasr, and serves as aux-
iliary in his forces.
After their acquisition of Sinope in 1324, the Candaroğulları became
much richer and more important, for that heavily fortified city had long
been established as a major port for the lucrative trade across the Black Sea.
Business, as usual, accompanied piracy. In 1340, for example, a small fleet
from Sinope captured Genoese and Venetian ships and were about to attack
another convoy when a Genoese admiral arrived on the scene, put together
an armed fleet, defeated the Turks, seized their lot, and executed their
crews.68 Nevertheless, Sinope remained a major port for international trade.
The emirate’s wealth is reflected in its coinage, the most abundant of
these beyliks. The issues of Suleyman Pasha (1309–c.1340) in the 1320s are
the most common, all struck in the name of the ruling Ilkhan, mostly of
distinctive types peculiar to this region.69 Surprisingly, though, he also
maintained close relations with the Mongols’ bitterest enemy, the Mamlukes.
The trade was evidently so important that it enabled him to pursue an inde-
pendent policy, perhaps playing his suzerains off against each other.
Unfortunately, the history of the emirate in this period is very poorly
known. Suleyman’s successors Ibrahim bey (c.1340–1345) and Adil bey
(c.1345–1357) struck coins in the name of the Ilkhan Abu Saìd, then of
Eretna, and anonymously in the 740s.
The Mongols
A unique accounting manual from the Ilkhanid period provides a conspec-

tus of the taxes and tribute paid by the various components of the empire.
After listing the revenues from Iran, Iraq, Azerbaijan and Georgia, the text
mentions the “central lands” (al-wastaniyya), the cities of Anatolia, of which
the westernmost are Ankara, Akşehir, Sivrihisar, and Qara Hisar.70 The
frontier lands (ucat) follow: Qaraman, awlad Hamid, Tughuzlu, Umurbeg,
Germiyan, Urkhan, Geredeboli, Kastamuniya, Akridur, and Sinub.71 These
are in geographical order except for Akridur, which is redundant and out of
place. Akridur was the capital of Hamid and then better known as
Felekabad.72 Sinop, though in correct geographical order, is another anom-
aly, for it was conquered by Kastamonu in 1324.
This frontier region stretches from the Mediterranean (Karaman) around
to the Black Sea (Sinope) and bears comparison with the emirates brought
into submission by emir Choban in 1315. Both comprise the Turcoman

emirates that were contiguous to the Mongol lands, but with significant
additions: Karaman, usually hostile to the Ilkhans, now appears as do two
states previously beyond the reach of the Ilkhans, listed here by the name of
their emirs, Umur bey of Aydın and Orhan the Ottoman. This is the first
and only record of Orhan paying tribute to a Mongol overlord. Naturally, it
would be important to determine the date of this document, which is
nowhere stated. It was presumably written when Umur had achieved fame
and wealth, probably between his conquest of Izmir in 1329 and his death
in 1348. Likewise, a date after 1326 when Orhan conquered his first major
city, Bursa, would be appropriate. On a larger scale, the work provides a
view of the empire at a time when it was still intact, so before the death of
Abu Saìd in 1335. The most probable date, then, would be the early 1330s.73
A Western source, written around 1332 confirms the general picture of
tribute to the Mongols: Turchi imperatori Tartarorum Persidis [the Ilkhan]
serviunt sub tributo, videtur quod Tartari deberent ipsos Turchos tanquam
suos contra nostros defendere.74
Meanwhile, Timurtash had resumed war against the Turcomans, now on
a vaster scale. In 1326, he took Beyşehir, capital of the Eşrefoğlu, executing
its bey and bringing the emirate to an end. He conquered the states of ’
“Aghizlu” with its important silver mines, “Tugancik” west of Trebizond and
Kirshehir. Ibn Sahib, the emir of Karahisar, saved himself and his domain by
taking refuge with Germiyan whose dependent he became.75 Farther west,
Timurtash took Sultan Önü and came to Philadelphia where, perhaps in alli-
ance with the Byzantine emperor, he forced Germiyan and Aydın to abandon
their attacks on the city.76 For a moment, Mongol power in Anatolia was at
its height, its domains as great as those of the Seljuks. But it was probably at
Philadelphia that Timurtash received the news of the fall of his family. He
decamped to Kayseri and from there took refuge in Egypt where the
Mamluke sultan eventually had him executed. After the fall of the Chobanids,
the Turcoman emirates regained their independence. Timurtash’s conquests
were ephemeral but, for the moment, the Turcomans were unable to gain
any territory from the Mongols, whose domain however was on the brink of
collapse.
The Turcoman Emirates
The rare abundance of contemporary sources provides considerable insight

into the nature and activities of these Turcomans. It shows first that the
confederations of nomadic tribes at the beginning of the century had in a

generation evolved into settled sophisticated states. High Islamic civiliza-
tion was everywhere evident in the presence of scholars, doctors of the law
and Koran readers who provided the rulers’ favorite entertainment. They
were rich, with luxurious palaces (Germiyan and Aydın), grand mosques
(Menteşe and Aydın) and entirely new capital cities (Menteşe, Karesi,
Ottoman). They were players on the international stage, having relations
with the Italian maritime states, Byzantium, the Mamluks of Egypt, and the
Golden Horde north of the Black Sea.
These relations included the major sources of their wealth: trade and the
raiding which brought booty and tribute. The maritime emirates and
Kastamonu profited especially from trade, but Menteşe and Aydın extracted
tribute from the Greek islands as the Ottomans did from Byzantium. Bribes
from the Byzantines to keep the peace or form alliances, and service as mer-
cenaries provided more ready cash.
All the states that had frontiers with Christian lands or were in easy reach
of them by sea practiced jihad but its aim was not religious but economic,
with justified wars against the infidel being a major source of revenue.
Despite their impressive and sometimes spectacular successes in war, the
beyliks suffered some setbacks when their opponents united. Constant
attacks—especially by Aydın—in the 1320s provoked a kind of mini-
crusade which in 1334 achieved a major victory, destroying the large fleet
commanded by Yahşi of Pergamum. The treaty Aydın signed in 1337
reflected its temporary weakness.
The beyliks differed in wealth and power: Menteşe, Aydın, and appar-
ently Kastamonu were the richest while Germiyan, Aydın, and the Ottomans
were the great powers of the day.
Most of these states were family operations, with the head of the family
based in the capital city and sons or brothers as governors of cities or prov-
inces. The most impressive example was Aydın, ruled by four brothers
under the suzerainty first of their father then of the eldest brother. For them,
division was not a source of weakness, for the family constantly cooperated.
Menteşe, Saruhan, Karesi, and Kastamonu had similar systems; the Ottomans
alone had only one leader (the situation in Germiyan is unclear).
In the 1330s, Aydın, thanks to the exploits of Umur Bey would have
seemed to be the leader among these states, though Germiyan was recognized
as the most powerful. Ibn Battuta was alone in appreciating the strength
and potential of the Ottomans whose conquest of Nicaea and Nicomedia in
this decade completed their control of rich and strategic Bithynia, laying the
foundations for an ever more powerful (and unified) state.
Notes
1. The Travels of Ibn Battuta, tr. H. A. R. Gibb, vol. II. Cambridge 1962.
2. Edited by Franz Taeschner: Al- Umari’s Bericht uber Anatolien in seinem
Werke Masalik alabsar fi mamalik alamsar, Leipzig 1929. Translated by
M. Quatremere in Notices et extraits des manuscrits de la bibliotheque du roi 13,
1838, 151–381.
3. Le destan d’Umur Pacha, ed. and tr. Irene Melikoff-Sayar. Paris 1954.
4. Texts and thorough commentary in Zachariadou 1983.
5. Francesco Balducci Pegolotti, La pratica della mercatura, ed. Allan Evans.
Cambridge MA 1936.
6. This section is basically a collection of raw material—a snapshot of a period—
which does not lend itself to being woven into a coherent narrative because of
the disparity of information about the various states.
7. I have omitted a few emirates that are not directly relevant to the present sub-
ject. The accounts also present much information about weights, measures and
prices which is not included here.
8. Described by Balban in al-Umari 32/350 (in this system of reference the first
number denotes the Taeschner edition, the second the Quatrmere translation),
but not included in his list of these emirates.
9. Text in Wittek 1934, 135ff.; RCEA 15, 5622.
10. See Vryonis 1971, 396–401; cf. “Ahi” by Fr. Taeschner.
11. Wittek 1934, 155.
12. “Fuka” of the text corrected to Mughla by Wittek 1934, 69.
13. Zachariadou 1983, 18–20, text 187–9.
14. Ibid. 159–73.
15. Ibid. 35f., text 195–200; f. 110 for the subordinate rulers.
16. Ibid. 60ff., text 217f.
18. Al-Umari 359.
19. Pegolotti 92, 104, 369.
20. Wittek 1934, 134–137; RCEA 14, 5591, 15, 5622.
21. The readings of Orkhan’s sole published coin (Ender 2000, 175) are quite
uncertain.
22. See Foss 2019, with further references.
23. Zachariadou 1983, 22, 27.
24. This may be an exaggerated reflection of Aydın’s now hostile relations with
Menteşe: Zachariadou 1983, 27.
25. Foss 1979, 149.
26. Pegolotti 43, 55f., 104, 369.
27. Pegolotti 57, 92, 104.
28. Zachariadou 1983, 36, text 190–4.
29. RCEA 15, 5657, 15, 5783, 5784.
30. For the date and circumstances, see Lemerle 1957, 28–39; Hızır may have
received his territories before 1326.
31. Ender 2000, 173f.
32. See Grousset 1970, 254.
33. Spuler 1955, 337.
34. For the location, at modern Bademye in the southern side of the Cayster valley
opposite Birgi, see Lemerle 1957, 35.
35. Lemerle 1957, 37.
36. RCEA 5657, discussed in chapter 4: “alamir alkebir almujahid almurabit abu
alkhairat sultan alghazat mubariz aldawla wal din.”
37. For what follows, see Lemerle 1957, 75–143 based on the Destan.
38. For Umur’s various attacks on Thrace, Greece, and the islands, see Lemerle 1957,
62–88, 102–15. After his meetings with Cantacuzene and Andronicus III in
1335, his efforts were directed against the Franks of Greece and Thrace, not
against Byzantine lands: Lemerle 1957, 116–79.
39. For the numbers and types of vessels, see Inalcık 1985, 205, 207f.
40. Inalcik 1985, 209–17 provides a useful analysis of Umur’s forces and motives.
41. Destan 97–1032, on which see Lemerle 1957, 105–15.
42. Destan 1033–84.
43. Pegolotti 289.
44. Uzunçarşılı 84–91.
45. Reported only in the Destan: Lemerle 1957, 63–74.
46. Lemerle 1957, 102–1115; cf Zachariadou 38f.
47. Cantacuzene II.65f., 77 ; Lemerle 1957, 148 n. 2.
48. The title also appears on their coins: Ender 2000, 23–7.
49. “Marmara” presumably represents the name of the sea of Marmora, indicating
that Yahsi’s emirate comprised the maritme parts of Karesi, which would accord
with other indications of the sources.
50. Pegolotti 43, 243, 369.
52. His coins identify him as “son of Timur,” i.e., of Temirhan: Ender 2000, 24.
53. Lemerle 1957, 96–9.
54. Gregoras I.538.
55. See Ender 2000, 23 (Demirhan, 1327–35), 24 (Suleyman), 25–7 (Yahsi
1327–43), and 28–35 (Beylerbeyi Celebi, 1343–?).
56. See Zhukov 1993.
57. Kafadar 1995, 116f.
58. Yaqub I: Ender 2005, 26f.; Mehmet: ibid., 30–7; Simav: ibid., 37.
59. Cantacuzene I.82; Zachariadou 1983, 28f. This was doubtless the occssion (or
one of many) for Byzanium paying a heavy bribe to Germyan.
60. See the Appendix to chapter 6.
61. See pp. 124, 126f, for the bribes paid to keep Orhan from attacking the region of
Nicomedia. The emperor paid an even larger tribute to the ruler of Germiyan:
see p. 209.
62. Cantacuzene I.505; Gregoras I.540; Zachariadou 64.
63. Beldiceanu 2000.
64. Toğan 1970, 318f.
65. See see above, p. 43.
66. Ender 2003, 53. The coin is anonymous, atributable by its style to the
Candarogullari
67. Cantacuzene III.284. The “Scythians” who lost Ankara in the same campaign,
were presumably the Eretnids, dominant in central Asia Minor at this time.
68. Heyd 1885, 552.
69. See Diler 2006, 476–80.
70. Risale-i Felekiyye 93a–93b.
71. Ibid., 93b.
72. Coins name Akridur in 699 and Felekabad from 707 (Izmirlier 1999, 109–113);
but Ibn Battuta 422 consistently refers to Akridur.
73. Remler 1985, 163f. suggests a date of c.1327.
74. Brocardus 503.
75. Balban in al-Umari 350; the names of these states have not been resolved.
Karahisar: ibid., 357.
76. Beldiceanu 1984, 22–9.
8
The Aftermath
The Turning Points
The decade of the 1340s saw a definitive shift in the balance of power among
the emirates. At the beginning, Aydin occupies center stage thanks to the
spectacular exploits of Umur, but ten years later, the Ottomans have become
the leading state. Sources for the 1340s are much poorer than those of the
preceding decade, leaving some states in complete obscurity but revealing
three events that held major significance for the future: the marriage of
Osman with the daughter of Cantacuzene in 1346; the death of Umur in
1348; and the Ottoman conquest of Karesi, apparently in 1349.
Menteşe appears to have passed the decade in peaceful prosperity by
remaining neutral in the wars in the Aegean and maintaining its trading
networks. Monumental construction continued, as attested by an inscription
(1344) of Orkhan’s son Ibrahim (c.1340–1360) on a mosque in Mughla.1
Aydın attracted the most attention thanks to Umur and the reaction he
stirred. His devastating attacks on Thrace, Macedonia, and Greece caused
considerable loss of population and provoked the western powers, urged on
by the pope, to create the Sancta Unio which captured the harbor fortress of
Smyrna in November 1344. It came to dominate the sea, destroying the fleet
of Aydın and Saruhan in 1347, but failed to win any victories inland. This
led to negotiations that made progress after Umur was killed in April 1348.
The truce signed in August 1348 showed Aydın’s weakness: it provided for
free trade, sharing customs revenue, and putting Aydın’s fleet in dry dock.
Aydın agreed to end attacks on Christians and undertook to protect them
from pirates of other emirates.2 The western advantage did not last long,
however, for the treaty made in April 1353 ended ten years of hostility
between Aydın—represented by Hızır and his brothers—and the Christian
powers. It dealt with trade and taxes and was far less unfavorable to Aydın.3
By then, Hızır had sought a counterbalance to Venetian influence in the
region by entering into friendly relations with Genoa which established a
consulate in Altoluogo soon after 1348. This naturally led to disturbed
DOI: 10.1093/oso/9780198865438.003.0009
relations with Venice, until a settlement was worked out in 1358.4 The emirate
recovered as western unity broke down, but, although Aydın could still fight,
it never regained the strength and prestige it had had while Umur was alive.
Philadelphia
A Christian island in a Muslim sea, Philadelphia survived a remarkably

long time, but constantly under threat. It may also have lost some of its
lands, for an inscription of his great-grandson claims that Mehmet bey of
Germiyan (1340–1361) conquered Guldi from the infidel. Guldi lies in the
volcanic hill country some forty kilometers north of Philadelphia and was
probably one of its outposts.5 In any case, Philadelphia was paying tribute,
presumably to Germiyan, in 1342.6
The city itself barely survived a determined attack in March 1348 by
Umur of Aydin together with his brothers—despite the friendship he had
established with Cantacuzene. The Turks even gained a part of the fortifica-
tions, but were pushed back. In the agreement that he made, Umur prom-
ised to leave the city in peace and left for Smyrna where he had to deal with
the Latins who occupied the maritime fort. He planned to renew the attack
on Philadelphia, but was killed in the fighting at Smyrna.7
Finally, in 1352, the Philadelphians, despairing of any hope from the
ever-weakening empire, sent a delegation to the pope in far-off Avignon.
They offered to submit their city permanently to papal authority. The pope
eventually gave an unhelpful reply that they should abandon schismatic
Orthodoxy and recognize the supremacy of the Roman church, adding that
the flames of hell were worse than anything the Turks could inflict.8
Despite the lack of support from outside, the enclave of Philadelphia
managed somehow to survive: an ecclesiastical document of 1369 described
the “holy metropolis of Philadelphia” as “preserved untaken till now, never
forced to bend its neck to any of the nations.”9 It survived until 1390 when it
succumbed to the Ottoman sultan Beyazit the Thunderbolt who conquered
the emirates as well.10
Saruhan
In 1345, the emir’s son together with the Suleyman of Karesi and Umur of
Aydin ravaged Thrace, this after Umur and Saruhan had resolved a territo-
rial dispute. Since Umur had lost control of the harbor of Smyrna, he could
The Aftermath 225
only send his forces to the Hellespont by crossing Saruhan’s lands. By mak-
ing a territorial concession, Umur established friendly relations with
Saruhan whose emir allowed his son to join the expedition.11 Also in 1345,
the rebel Vatatzes had requested help from Saruhan, but met with a refusal.12
Finally, in 1347, Saruhan joined Aydın in putting together a fleet of 118
small vessels, only to be destroyed by crusaders off the island of Imbros.13
Karesi
In 1341 and 1345 Yahsi attacked Christian lands together with Saruhan; its
contingent in the second expedition led by Temirhan’s son Suleiman. The
first was badly defeated; the second inflicted real damage on Thrace.14 In
1345, During the civil war between Cantacuzene and the dowager empress
Anna, a certain John Vatatzes, thanks to a close alliance with Suleyman,
“satrap of Troy” who had married his daughter, had no trouble attacking
Thrace for the empress; but the Turks, who had had established good rela-
tions with Cantacuzene, soon switched sides and killed Vatatzes.15 Likewise,
when the panhypersebastos Isaac Asan, a partisan of the empress, sought help
from Suleyman against Cantacuzene, his large bribe was refused and when a
subsequent official arrived on the same mission he was shown the door.16
This the last appearance of Karesi in a Byzantine source; the Ottoman
version is very different. Aşıkpaşazade (APZ) begins by mentioning “Aclan
bey” son of Karesi who has two sons “Dursun bey” and “Haci İlbeg.”
Dursun, who takes refuge with Orhan, proposes to turn Balıkesir, Bergama
and Edremit over to him, retaining only two small places on the coast west
of Edremit—divisions that correspond to no known Karesi reality. Nor is it
obvious how the names of the rulers can be reconciled with those known
from other sources (though see next). Note that Byzantine sources make no
mention at all of the fall of Karesi, though it was Orhan’s acquisition of this
coast that made the momentous crossing into Europe possible. An early
Ottoman source, the Chronological List of 1421, dates the conquest of Karesi
to AH 749 (=1348/9).17 Although this seems plausible, the circumstances
remain totally obscure.
Germiyan
Although Balban mentions Germiyan as taking territory from its neighbors

there is little evidence of major expansion. Only a very long and detailed
inscription of 1411, known as the taş vakfiye or “stone endowment” gives an

example. In it, the current ruler of Germiyan records that his great-
grandfather Mehmet bey (1340–1361) took Guldi and the lake of Simav
from the infidel.18
Although Guldi may have been in the territory of Philadelphia, lake
Simav is in Phrygia, much farther to the north.19 It is hard to imagine any
infidel power that could have held Simav at this time. Most probably,
Germiyan took Simav from Karesi, but didn’t want to flaunt a conquest
from a fellow-Muslim. Curiously, the only coin of Mehmet that bears a
mintmark was apparently struck in Simav.20 On the other hand, taking land
from Philadelphia was consistent with the attacks that Germiyan had been
making since at least 1304.21
Ottoman
With their capture of Nicomedia in 1338, the Ottomans completed their

conquest of Bithynia and established themselves as a major power with
an urban base and posing a greater threat than ever to the declining
Byzantine empire. Their aggression was now directed westward. They
made no move to the east, where the small emirate of Gerede continued
to exist, nor to the south where they faced Germiyan. Their advance was
directed not against Byzantium, but the neighboring emirate of Karesi
which they conquered under obscure circumstances by the end of the
decade. This completely transformed the balance of power in western
Asia Minor, where Orhan now controlled a vast region from the Sangarius
to the Aegean, with the coasts of the Marmora and the Dardanelles. No
other emirate had such potential or strategic location, especially once
Umur was removed from the scene. Byzantium was quick to recognize
the situation. John Cantacuzene, who had been proclaimed emperor in
1341 but still had to finish a civil war to gain complete supremacy, had no
doubt of Orhan’s dominance among the emirates. Therefore, when a
request arrived from Orhan for the hand of his daughter in marriage, he
agreed and in 1346, a Turkish emir became son-in-law of a Byzantine
emperor. The alliance thus formed made it natural for Cantacuzene to
bring an Ottoman force across the Dardanelles to Europe, with earth-
shaking consequences.
The Aftermath 227
Kastamonu
The Candaroğulları remained active in international trade: the Genoese had

a consulate at Sinope by 1351 as did the Venetians at an uncertain date.22
But their history is poorly known. In the 1340s the emirate struck coins in
the name of Eretna or anonymously from Sinope, Kastamonu, Burghulu,
and possibly Bolu and Gerede.23 The unique coin of Gerede, discussed ear-
lier, may indicate that their territories had expanded to the west, making
them a direct neighbor of the Ottomans.
The Mongols
When Abu Saìd died in 1335 without an heir, the Ilkhanate sank into civil
wars that resulted in Anatolia breaking away to form an independent state
called the Eretnid after its founder.
Alaeddin Eretna rose as a follower of Timurtash who made him gover-
nor of Anatolia. Ibn Battuta, who met him in Sivas, called him “the lieu-
tenant of the king of Iraq in the land of Rum,” noting that he spoke
educated Arabic, and was generous. One of his wives resided at Kayseri,
the state’s main military base, while a deputy administered Aksaray. At this
time Sivas was the largest city of Anatolia and base of the civil government,
while the army headquarters was Kayseri. The Ilkhanids ruled from
Baghdad and Soltaniyeh in Azerbaijan, and from the mahalle, their vast
traveling camp.24
At the death of Abu Sa’id, Hasan Bozorg (“Big Hasan” head of the Mongol
Jelayrid tribe), the governor of Anatolia, defeated his rivals, and promoted
Eretna to be governor. These years were marked by the factional rivalries
that ultimately destroyed the integrity of the Ilkhanid state. In 1338 Kuchek
(“Little”) Hasan the son of Timurtash, defeated Hasan Bozorg, installed
Sulayman as Ilkhan, and struck coins in his name from 739 to 746. Eretna
exploited the factional divisions, but when Hasan Kuchek started to move
into Anatolia, Eretna appealed to the Mamluke sultan Nasir who confirmed
him as governor of Anatolia. He struck coins from his capital Sivas in name
of Nasir 739–741.25 In 1343 he solidified his independence by defeating his
greatest rivals Hasan Kuchek and Suleyman at Karanbük. Eretna’s abundant
coinage, struck in mints throughout Anatolia from Erzurum to Ankara in
742–753/1343–1352, reflect his domination of Anatolia, independent of the

Ilkhanids. They are all anonymous but bear the title Sultan Adil in Uighur.26
The history of his state is very poorly known. Eretna was aware of events in
western Anatolia but unable to influence them. In 1344, when news arrived
of the capture of the harbor of Smyrna by the crusaders, he sent not an army
but two engineers skilled in making catapults.27
Eretna had a reputation for justice and piety. His death in 1352 was fol-
lowed by another period of factional/dynastic wars, which permanently
drained the power of the state.
Notes
1. Wittek 1934, 137f.; RCEA 15, 5983.

2. Zachariadou 1983, 55f.
3. Ibid. 54–60; text 211–16.
4. Foss 1979, 154f.
5. The extent of Philadelphia’s territory is unknown; only the mention of Fort St.
Nicholas indicates that it extended to the headwaters of the Cogamis which
flowed through Philadelphia and eventually into the Hermus.
6. Miklosich and Muller I p. 228; for the meaning of the term employed, verimion,
see Zachariadou 1983, 24.
7. Lemerle 1984, 55–67, quoting a Greek text that gives considerable detail about
the fortifications.
8. Schreiner 1969, 401f.
9. Text and translation: Foss 1976, 125–8.
10. Schreiner 1969, 404–11.
11. Cantacuzene II.529f.; Lemerle 1957, 204–17.
12. Cantacuzene II.553; Lemerle 1957, 219f.
13. Zachariadou 53.
14. Cantacuzene II.65–70, 77, II.529–34; cf. Lemerle 1957, 148n.2, 204.
15. Lemerle 1957, 219f., who deals with contradictions in the sources.
17. Osmanlı Tarihine Ait Takvimler, ed. C. N. Atsız (Istanbul 1961) 24f., no. 136.
Note that this calendar gives the correct dates for the conquest of Bursa and
of Iznik.
18. Varlık 147, lines 5, 6; available online in transcription: https://tr.wikipedia.org/
wiki/Taş_Vakfiye
19. Foss 1985, 100–2.
20. Ender 2005, 37; the reading is uncertain.
21. See 177f.
The Aftermath 229
22. Heyd 1885, 552.

23. Ender 2003, 50–84.
24. Ibn Battuta 335–44.
25. Perk and Öztürk nos. 456–9.
26. Perk and Öztürk 191–276.
27. Destan 2091f.
9
Final Thoughts
This concluding section is deliberately not called “Conclusions” because it is

not yet possible to close the book on the earliest Ottomans: too many pages are
missing. Knowledge of Osman’s career depends on a very dubious tradition:
you can take it or leave it or—much more often—pick and choose the bits that
seem to make sense. Osman bursts on the international scene in 1302 and
has five years of attack, leading his faithful nomads/powerful army against
the heavily fortified bastions of Byzantium. He is on the verge of spectacular
success when he’s struck down by the Boss—the Ilkhan—and drops out of
history. When the curtain rises almost twenty years later, the same scenes
are there; nothing has been moved but Orhan, who will almost become a
three-dimensional character, moves inexorably to victory, capturing the
Bithynian cities one by one in a decade. Yet he has inherited something
from his shadowy father: a sophisticated Islamic apparatus which puts him
on a par with the other Anatolian emirs. Before we lose sight of him as he
crosses to Europe, he has left his neighbors behind and is in the forefront of
conquest from the now feeble infidel empire and its neighbors and allies.
The problems of reconstructing the beginnings of this history are so great
that the leading English scholar of these years has issued a firm and plausible
warning:
The best thing a modern historian can do is to admit frankly that the earliest
history of the Ottomans is a black hole.1
Black hole indeed, but occasional glimmers of light manage to escape, as I

hope these chapters have shown.
Investigation of the Ottoman Homeland presents a real landscape and
confirms the existence of virtually all the places mentioned by APZ. They
are not only in the right place but at the right time, for most of them have
remains dating from the thirteenth-fourteenth centuries, notably the network
of fortresses built by the Byzantine Lascarids. The authors of the tradition
knew the country very well, for their work gives occasional details that
would only be known to someone familiar with the region.
DOI: 10.1093/oso/9780198865438.003.0010
The Homeland is a rough country of narrow valleys and low, often steep
mountains, a transitional area between the central Anatolian plateau and
the rich maritime region to the north. Unlike those, it had no ancient cities
but was until modern times marked by small settlements of limited
resources. It was not ideally suited to the large flocks of nomads but had the
advantage of lying on important routes that connected the coastal region
with the interior, making trade and predation possible economic activities.
The narrow valleys and steep hills of the Homeland precluded pastoral-
ism on the Ilkhanid scale which demanded large open plains for their vast
flocks and herds, as Osman’s grandfather Suleyman is reputed to have said
after spending some years in Anatolia: “the mountains and valleys of Rum
caused them damage, for the nomads’ sheep suffered from the valleys and
the peaks.”2 Nevertheless, the tradition has Ertuğrul and his 400 tents mak-
ing their unceasing migrations between summer and winter pastures.
The Ottomans were certainly practicing transhumance in Orhan’s time,
for the governor of Mesothynia warned the emperor before the battle of
Pelekanon to attack the Turks before they withdrew to higher ground in
their annual migration.3 Likewise the emir of Aydın had gone up to the cool
mountains when Ibn Battuta arrived.4 Some, at least, of the Turks beyond the
Sangarius were evidently nomads, for Michael Palaeologus in 1281 found
only their abandoned campsites and in 1302 Ales Amourios requested land
from the emperor where he could settle his forces permanently—implying a
group constantly on the move like the tribes who were breaking though the
imperial frontiers at that time.5
In other words, the early Ottoman polity included an element of pasto-
ralism, but the landscape makes it unlikely that nomadism was a foundation
of their state. To judge that it would be necessary to know the balance
between pastoralism and agriculture, but of the latter the sources reveal
nothing until the Ottomans had moved into the fertile lands of maritime
Bithynia.6
A real environment needs to be inhabited by real people; the fact that the
environment was real doesn’t mean that any of the events recounted in the
tradition actually happened or that the people whom they portray lived
there or anywhere else. In fact, the tradition contains so much confusion,
downright mistakes, or impossibilities that a careful reading leads to skepti-
cism or rejection. It seems clear that the “information” it gives has passed
through a stage of unstable oral transmission, with many edifying stories or
folktales incorporated in it. Getting at the actual events and development of
Osman’s early career (before 1302 when corroborating Byzantine sources
are available) seems impossible.
Final Thoughts 233
Some developments, though, seem plausible, whatever the details may

have been. Ertuğrul may be no more real than Romulus but the idea that the
ancestors of the Ottomans settled in Söğüt has never been open to doubt,
and might as well be accepted, along with the notion that a sovereign
assigned them land in a frontier district where Seljuks had faced Byzantines.
That sovereign need not have been the Ala ed-din around whom legends
grew. Likewise, the gradual expansion of Osman’s territories toward the
Byzantine north makes sense in the light of subsequent events. For what-
ever reason, these Turks were directing their efforts to lands held by the
preeminent Christian power. In any case, they could not advance southward
(even if they wanted to), for there they faced the powerful Germiyanids and
smaller states which set Osman’s bounds at Inönü and Eskişehir—to say
nothing of the Mongols who lurked on the nearby Anatolian plateau. In
general, Osman did not attack territories held by Islam (the confusing case
of Karacaşehir seems impossible to resolve; traditions about it may be alto-
gether apocryphal) The lands to the east seemed good for raiding but not
conquest for they offered none of the wealth of metropolitan Bithynia, with
the imposing cities of Prusa, Nicaea, and Nicomedia.
The Homeland has an unpromising characteristic: it is a narrow land
with limited resources, not capable of supporting the large forces needed to
take on Byzantium. For that, Osman needed support or collaboration of the
kind the tradition supplies in the figures of Köse Mihal, Samsa Çavuş, and
others—all of them probably mythical. Yet, he appeared, seemingly out of
nowhere, with a substantial army near Nicomedia in 1302.
Significantly, Pachymeres relates that Osman was joined by forces, most
probably Turcoman, from Paphlagonia and the Maeander. If they joined
him, it was presumably because he had already established a reputation,
most likely as a fighter against the infidel, a ghazi. There is no reason to
suppose that he had recruited many from the local Christian population or
that he represented a blending of Turks and Greeks. On the other hand,
some defenders of the Byzantine frontier and its fortresses, disgruntled by
the incompetence, avarice, or neglect of the government, did switch sides,
often enough to weaken the frontier defenses.7
Although the tradition is plausible in general outline—Osman, starting
from Söğüt, gradually conquered territory from “tekfurs,” moving north
from poor mountainous inner Bithynia toward the rich lands of the mari-
time district where he confronted powerfully fortified cities—it cannot be
relied on in detail, for it is full of distortions, mistakes, and myths. They
start with the original lands, where summer and winter pastures are at the
same elevation and separated by hostile territory, include legendary figures
like Köse Mihal, folktales, and significant dreams. Curiously enough, they
don’t get better as the narrative moves into a historical period when it can
be checked against other sources. The errors and inconsistencies of the
tradition are striking: it includes events like the battle of Dimboz, unattested
elsewhere, but omits Osman’s victory at Bapheus.
In the account of Orhan’s time, the tradition confuses the conquest of
Iznik and Izmit and—most surprising of all—omits Pelekanon where the
Turks defeated the Byzantine emperor in person. It cannot clarify the con-
fusion attendant on Karacahisar. Even the conquest of Karesi is inconsistent
with other sources. As a result, it is not possible to rely on the tradition for
the origins of the Ottoman state or its development under Orhan.
On some points, though, the tradition may make a real contribution. Its
accounts of relations with the sultan Ala ed-din, who supposedly granted
land to Ertuğrul and rewarded Osman for his conquest of Karacahisar look
very dubious. But Osman had a laqab or honorary name which would nor-
mally be granted by a higher authority such as the sultan. Likewise, the
selection of Yenişehir as Osman’s capital (attested only by the tradition)
makes perfect sense. He chose a site well located for moving against the
cities and trade routes of inner Bithynia. Yenişehir is only some twenty
kilometers from Iznik (though on the other side of a mountain) and in
striking distance of relatively easy routes to Bursa in the west and the
Sangarius on the east. Establishing his new capital there would have left no
doubt as to his intentions, and would account for his sudden appearance
outside Nicomedia. Finally, though the figures of Köse Mihal, Samsa Çavuş,
and the others may be entirely legendary, they could represent the rein-
forcement Osman was getting from other tribes at a time when they were
swarming over the Byzantine lands when the frontiers had collapsed, a time
of flexible loyalties, as Cantacuzene remarked (apropos of a slightly later
period): “it is the custom among these babarians, when one of them goes on
a campaign, that those of another satrapy who want to join it, are not
pushed aside but received with pleasure as allies.”8
Starting from a small, poor, mountainous, landlocked Homeland,
Osman’s resources were limited: he controlled no major cities, struck no
coins, built no surviving mosques, set up no inscriptions, and made grants
only on a small scale. Big changes came with the establishment of Yenişehir.
By 1304 he had advanced down the Sangarius where he took Malagina, the
dominant fortress of the region which opened for him the routes to
Nicomedia, the interior, and the Black Sea. In the next two years, he was
putting Nicaea and Prusa under blockade. Osman was making tremendous
progress until the Mongols, called in by Byzantium, slapped him down so
Final Thoughts 235
effectively that he disappears from the record for close on twenty years. In
defeat, he lost all his recent conquests and was pushed back to the moun-
tains above Bursa—presumably to the Homeland, and here history leaves
him. Yet, there is evidence to suggest that he made some recovery for the
document endowing the hospice of Mekece in 1324 strongly implies that by
that date the major roads along the Sangarius and between Nicomedia and
inner Anatolia were pacified, presumably the work of Osman preparatory to
renewed onslaught on the Bithynian cities where the returning light of his-
tory finds the forces of Orhan, poised for the conquests that would make
the Ottoman a major regional power.
Everything changes under Orhan, to such an extent that it would seem
logical to call the emerging state the Orhanian rather than the Ottoman
Empire. He is hardly in power when Bursa falls to his blockade and becomes
his capital; five years later it is the turn of Nicaea and seven years after that
Nicomedia. Orhan is now master of the richest and most strategic part of
Bithynia and the unwelcome neighbor of Byzantium. He sends raiding
expeditions into the districts he doesn’t control directly. He is the master of
major trade routes and a strategic seacoast. He strikes coins, sets up com-
memorative inscriptions, and builds on such a scale that he completely
transforms his new capital into a substantial Islamic city. He produces doc-
uments that show a high degree of sophistication and has scholars and
Koran readers present in his court. Note that the document that attests high
Islamic civilization was necessarily issued from the obscure Yenişehir, not
Bursa which had not yet been conquered.
The transformation from Osman’s backwoods domain into Orhan’s met-
ropolitan principate is sudden and dramatic, even more so when viewed
from the 1340s that saw the emir married to the emperor’s daughter and his
lands stretching to the Dardanelles and the Aegean by the conquest of
Karesi. Even bigger changes followed the crossing of his troops into
Europe—but that is beyond the present the subject.
All this cries out for an explanation, and the past century has produced
many, much like efforts to explain the Fall of Rome—and with equal
success:
‘Tis with our judgments as our watches, none

Go just alike, yet each believes his own.9
The debate, which has long been raging, need not be followed in detail here
since one of its participants, Heath Lowry, has produced an excellent account
of its various theories; a brief summary of the most important will suffice.10
Herbert Adams Gibbons, an American journalist stationed in Istanbul,

opened the discussion with his The Foundation of the Ottoman Empire in
1916. He proposed that the Ottomans were a new race, a blend of Turks and
Christian converts, that they inherited Byzantine administrative practices,
and that this amalgamation of peoples and traditions accounted for their
success Although his main points find no support now, his idea that
Ottoman society was in some way special can hardly be denied considering
its remarkable success.
In 1924, the renowned Turkish historian Fuat Köprülü successfully
showed that Ottoman institutions had an Islamic origin and in 1934 went
one step further by maintaining that the Ottoman state was purely Turkish
and that it represented the continuity of the Oğuz tribe, a notion that
became a Turkish orthodoxy.
Two Harvard professors, William Langer and Robert Blake, in 1932
introduced a new element by stressing the importance of heterodox reli-
gious movements that made it easier for Christians to convert, and pointed
out the fortunate circumstance that the Ottomans were neighbors of the
ever-weakening Byzantine empire. They introduced geography, sociology,
and trade into the debate.
The most controversial theory, whose discussion has dominated the
debate since it was advanced in 1937, was Paul Wittek’s characterization of
the Ottoman as a “ghazi” state, that is, one whose raison d’être was the holy
war against the infidel, a community of warriors united in their determina-
tion to advance the cause of Islam.
A potentially important contribution was made by Zeki Velidi Togan in
his Introduction to the General Turkish History (1946) where he put the
Ottomans and the Beyliks into the context of Mongol Asia Minor—and
potentially important because this work has been rarely been noticed or
referenced.11 He noted the enthusiasm without fanaticism of the early
Ottomans, along with the weakness of Byzantium and the significance of
the Homeland’s location on a major trade route.
A Greek professor, G. G. Arnakis, in 1947 produced a comprehensive
account of the early Ottomans that has no rival but is rarely cited because it
a written in Modern Greek. He investigated conditions in Bithynia with its
disaffected population and followed the course of Ottoman expansion, con-
fronting Greek and Turkish sources at every opportunity. He regarded
intermarriage and conversion as a source of Ottoman success by increasing
the available manpower and saw the Akhi brotherhoods as a factor in main-
taining social stability.
Final Thoughts 237
Beginning in 1967, Irene Beldiceanu-Steinherr published articles and

monographs that revolutionized study of this field by exploiting documents
from the Turkish archives. These gave her unparalleled insights into the
society and economy of the early Ottomans.
Turkey’s leading Ottoman historian, Halil Inalcık, in 1982 (and other
publications) portrayed the Ottomans as a cosmopolitan state, incorporat-
ing many Christians into a society of Turkoman tribes often attracted by the
prospect of holy war. Inalcık, like many others, holds up the figure of Köse
Mihal as an example of Christian–Muslim collaboration.
A new approach inspired the 1983 work of Rudi Lindner who applied the
methods of anthropology to the structure of Ottoman society, positing a
tribal, inclusive image that saw nomadism as a central feature of Ottoman
life. He expanded on this thesis in his 2003 monograph.
Cemal Kafadar of Harvard in 1999 took an inclusive attitude, incorporat-
ing elements of the ghazi theory with epic accounts of Turkish tribal origins,
but did not advance a new general interpretation.
In 2003 Heath Lowry saw a “frontier society in flux” with a place for
everyone—Muslim or Christian, free or slave—in a complex and sophisti-
cated society, with elements of high Islam and Seljuk administrative prac-
tices. He explained gaza as meaning raids for loot as well as holy war.
None of these explanations has gained universal approval. It is not my
intention to add to them, but simply to point out some factors that might be
incorporated into the debate.
Although Lowry’s seems the most plausible of these, one element is
totally lacking from his exposition, and rarely stressed by any of the others.
That is the Mongols, to whom this discussion will return.
Osman started with a rough unpromising homeland that no one would
covet. It could not support a large population, but Osman needed man-
power if he were to move successfully against Byzantium. How, then, to
attract them? The prospects of loot—in this case from regions near the
Byzantine capital—was no doubt a factor, as it had traditionally been for the
nomadic rulers of the vast Asian steppe. In their case, war for booty was a
central element of their power, necessary to justify their rule over inde
pendent tribes.12 Yet the idea of advancing the frontiers of Islam against a
rich and well-established infidel power would also have proved attractive.
This does not mean that Osman’s state was organized around the holy war,
but simply that motives for fighting in Bithynia could have been more com-
plex than simply loot. Islam had a long tradition of fighting the jihad, a
cause embraced in Osman’s time by his fellow emirs in Asia Minor, on the
frontiers of Islam. Another source of manpower was probably the product

of conversion of the indigenous populations to Islam or at least intermar-
riage with the Turks. Several of the scholars discussed earlier favor this
interpretation, for which documentary evidence has been adduced, show-
ing that the population of Bithynia was largely Turkish by the end of the fif-
teenth century and perhaps already by the fourteenth.13
After establishing Yenişehir, Osman faced an empire in accelerating
decline. Byzantium was reeling from what seems to have been a sudden col-
lapse of its eastern frontiers where newly arrived tribes were moving
around, not yet settled. Osman had real advantages. He confronted an
empire that had to defend a vast area and naturally gave priority to Ionia,
much richer than Bithynia; but after 1304 that effort was doomed, leaving
Byzantium with only a fraction of its former Anatolian realm, already under
attack from Osman. Since his main effort was directed against a Christian
power, he could call on ghazis. He constantly expanded his base at the
expense of Byzantium, attacking Christian lands and raiding those (as in
eastern Bithynia) under Muslim rule. His administration (at least in its last
years), as the Mekece vakf suggests, was sophisticated and infused with
orthodox Islam; he had the manpower to attack walled cities, leaving
detachments to blockade them through long years.
At the crucial time when Osman’s rise began, he was fortunate in the pre-
occupations of greater nearby powers. Germiyan was involved in a long
struggle with the Sultanate, from 1286 to 1289, just as Osman was begin-
ning to expand. More advantageous, perhaps, were the revolts that plagued
the Ilkhanids: the devastation inflicted on southwest Anatolia by Gaikhatu
1291–1292, the contemporary revolt of Paphlagonia only suppressed in
1293, revolts in 1296 and 1297, and the massive uprising of Sulemish
1298–1299. These would all have tied the hands of the Mongols, making it
difficult if not impossible to control the largely autonomous tribes of the
frontier, especially since the main aim of Ghazan, who restored control in
Anatolia, was not the west at all, but Syria where he led unsuccessful expe-
ditions in 1299–1301 and 1303. He was occupied with these wars just at the
moment when Osman appeared on the scene. It was not until 1307 that the
Mongols could turn their attention westward, with near fatal results for
Osman, who had been doing very well in the meantime.
When Orhan came to power, his state was notably poorer than Aydın or
Menteşe, who had numerous rich fortified cities and seaports, who profited
from extensive trade with the west. That trade, supplemented by raiding and
providing mercenaries, generated the wealth manifest in their grand
Final Thoughts 239
mosques and new or transformed capitals. Saruhan functioned the same

way, though evidently on a smaller scale. Karesi was involved in the same
activities and was rich enough to build a new capital. Kastamonu was thriving
after the capture of Sinope and profited from the Black Sea trade. As late as
the 1330s, thanks to the exploits of Umur, Aydın could have been considered
the leader among the emirates, but the conquests of Orhan were beginning
to tip the scales in his favor.
Like his father, Orhan profited from circumstances. Most decisive was
the decline of Byzantium, weakened by civil wars (1320–1328 and
1341–1347), constant attacks from the west, and the loss of its rich lands in
Ionia. He built for permanence, increasing his power as he captured the
Bithynian cities one by one. Here he was very different from Aydın, where
Umur was interested only in loot, as Menteşe apparently was with business.
Geographically and politicly, Orhan was in the best position to expand
since he faced an empire that grew weaker as he conquered its lands. Neither
Aydın nor Menteşe was well situated for conquest, for they had to face the
Italian states, the knights of Rhodes, and other Christian powers who could,
when necessary, unite against them. Karesi would seem to have an advan-
tage, controlling a long seacoast with the western Marmara, the Dardanelles,
and the northeast Aegean, yet it played a minor role, perhaps because its
rulers were divided, with the senior favoring Byzantium and the subordi-
nate attacking it. Karesi provided mercenaries and unreliable alliances, it
seems, generally on a small scale, but this emirate is very poorly known.
Finally, Kastamonu controlled a large territory, most of it mountainous and
difficult of access but astride major trade routes that led to the Black Sea. Its
location, though, was too far out of the way to play a role in the events con-
sidered here.
Because Orhan was confronting the greatest of the infidel powers, he
could no doubt attract other warriors or tribes, ghazis anxious to join the
fight if not for the Faith, at least for loot. But so did all the others, as attested
by their inscriptions and actions. Umur was qualified as ghazi and Aydın
practiced jihad, but none of them achieved any great enduring success like
the Ottomans whose domains constantly expanded as they incorporated
the lands they took. They could administer an ever-growing state because
they had the apparatus of high Islamic civilization with its sophisticated
chancery, its eunuchs, and slaves. But that, too does not distinguish them
from the rest, who were also part of a high Islamic tradition. Again,
Ottoman determination and their providential location gave them the
advantage. Also important was the fact of their efforts being directed against
a Christian power; they raided Muslimi to the east, but for a long time made
no effort to conquer them. At least in theory, this should have made the
Ottomans most appealing to adventurers and fighters eager to confront the
infidel. But such decisions were probably made on material rather than
ideological grounds. Here, too, the conquests which brought loot and
increased revenue would have made the Ottoman state increasingly rich as
its reputation presumably grew with its success. If any of the maritime emir-
ates wanted to expand, they would have to fight fellow Muslims; the case of
Kastamonu was similar.
Another factor probably contributed to Ottoman success. They main-
tained a consistent unity of rule, with only one supreme leader at a time.14
In this, they differed from the other emirates where brothers or other close
relatives shared power. Aydın was the most notable example, with four
brothers ruling their own districts—and cooperating. But in Karesi two
brothers followed very different policies, pro- and anti-Byzantine, which
may have been a factor in their ultimate collapse. In any case, Osman and
Orhan (and their immediate successors) ruled alone. APZ (cap. 29) reports
that Orhan had a brother Alaeddin who, when Osman died, told Orhan that
it was best to have a single ruler, and retired to a farm in the vicinity of
Bursa.15 This incident may be apocryphal, but Orhan had a better attested
brother, Pazarlu, who commanded part of the Ottoman army under Orhan
in the battle of Pelekanon (1329)—i.e., as a subordinate not an equal.
Although Orhan had at least five brothers, he succeeded Osman without a
colleague.16
Being in the right place at the right time doesn’t sound like much of an
explanation, not even when combined with an unusual determination to
move forward, to conquer and to take over new lands, so that the wealth
and power of the state increased as it conquered. A similar explanation
could as well fit the rise of Rome. But in that case, ethnic questions arise, for
it was not only Romans who did the conquering but former allies and ene-
mies like Latins, Sabines, Etruscans, and others who were eventually
absorbed into the body politic and became Romans. Ottoman success ulti-
mately depended on manpower and here the theories that see Christians—
whether converted or not—as part of the dynamic would have value,
perhaps creating not a new blended society, but at least one where the
locals could join in the common effort, usually through conversion or
intermarriage.
So far, this discussion has looked outwards, toward Bithynia and
Byzantium, but there may be lessons to be learned in looking inward, to the
Final Thoughts 241
powers that were for long far greater than the Ottoman. These are Germiyan
and the Mongols—not the Seljuks who had effectively lost the reality,
though not the appearance, of power by the time our narrative begins.
Relations with Germiyan have already been considered. The tradition
minimizes this powerful state, presenting it as a troublesome neighbor,
nothing more. In fact, it was the dominant beylik of western Anatolia, exer-
cising control over Sasa bey, Aydın, and Menteşe, if not others. It is not pos-
sible, however, to gain any insight into its relations with the Ottomans in
this period.
However powerful in reality, Germiyan could not be compared with the
Mongols to whom it came to offer submission in 1316, and to whom it was
paying tribute in the 1330s. The same document lists Orhan as a tribute
payer, a fact that the tradition doesn’t even hint at. In fact, the Mongols are
almost completely absent from the Ottoman sources which only mention
the Çavdar Tatars as disturbing the peace, and casually the Tatar Bayıncar as
destroying an Anatolian city, Ereğli (APZ 6).17 This absence is reflected in
the modern accounts, from Gibbons for whom “the Mongols were never
more than mere raiders in Asia Minor” (p. 37) to Lowry who doesn’t men-
tion them at all. Most of the others treat them in passing. They are central
only for Togan and Lindner.18
Yet there are a few sources that show the presence of the Ilkhanids was
overwhelming, beginning with the will of ibn Jaja which portrays a settled
and prosperous region around Eskişehir flourishing under their rule in the
1260s and 1270s. That means that the Mongols were direct neighbors of
Ertuğrul and Osman, who grew up on the doorstep of the descendants of
Genghis Khan. In 1307, Osman was forcibly reminded of their power—if he
needed reminding—when they defeated him and drove him back to the
Homeland, abandoning most of his conquests. Coins of the emirates that
issued them shortly after 1300 reflect a real or symbolic Ilkhanid superiority
in Germiyan, Karesi, Kastamonu, Aydın, and the realm of Sasa. Even the
first coins of Orhan are in an Ilkhanid style, though they bear Orhan’s name
alone. The expedition of Choban in 1316 received the submission of the
emirs of the frontier and the tax treatise shows they (including Orhan) were
tribute payers in the 1330s. There seems to have been a hierarchy in which
Germiyan dominated the other emirates but was itself subordinate to the
Ilkhan. The Mongols themselves rarely intervened in the western emirates,
perhaps counting on Germiyan to ensure stability. It is probably not a coin-
cidence that Orhan’s greatest successes began after the Ilkhanids were
embroiled in internal dissention, then collapse in 1335. Everything suggests
that the Mongols were a prime factor in the life of the growing Ottoman
state and that future research should take them into account in a serious way.
To sum up, there is no one factor that can explain the rise and success of
Osman and Orhan, but several that contributed.
Perhaps most importantly, they were in the right place at the right time:
the Homeland was on the frontier of Byzantium and particularly the rich
plains and cities of maritime Bithynia, in proximity to the capital. If Osman
were to expand, he would necessarily move north against the infidel, for his
southern and eastern flanks were blocked by the greater powers of Germiyan
and the Ilkhans. The Homeland was a rough and relatively poor area, of no
interest to other emirates which never made any effort to conquer it.
Overcoming its natural disadvantages may have served as a stimulus to
expansion. Its lack of a large population could have inclined Osman to seek
allies (represented in the tradition by various semi-mythical figures) and
make use of the local Christian population. He may have possessed some
sort of charisma or leadership qualities that encouraged people to follow him.
These last points are pure speculation and don’t explain Osman’s dyna-
mism after decades of Ertuğrul’s inactivity. The sudden collapse of the
Byzantine frontiers (cause unknown) around 1300 certainly provided an
occasion for movement.
There are two ways in which Osman differed from his fellow emirs: he
and Orhan ruled as individuals, not as part of a family sharing power; and
he was no mere raider, campaigning for loot then returning home, but he
annexed territory as he conquered it, thus growing stronger at every stage.
As for the right time, he appeared on the international scene at a moment
when Byzantium was embroiled in a war with Venice, and generally faced
an empire beset by hostile powers: the Italian and Greek states, Serbia, and
Bulgaria were chronic ever-present enemies, some so close to home that
they had to be fought or placated whatever the situation in Asia Minor.
Conflict within the Church and frequent palace conspiracies didn’t add to
the security of Byzantium and further weakened its ability to defend its
eastern frontiers.19 Likewise, the potential enemies in Osman’s rear—
Germiyan and the Mongols also had other problems to face: Germiyan at
war with the Sultanate when Osman’s career was beginning and Ilkhanid
Anatolia plagued by revolts when he was expanding. Orhan had even
greater advantages—the Byzantine civil war and the collapse of the
Ilkhanids in 1335—just as he was completing the conquest of Bithynia. His
acquisition of the Bithynian cities vastly increased his wealth and power,
laying the foundation for expansion into Europe and conquests beyond the
scope of this work.
Final Thoughts 243
These are some factors worth considering, but they are not the whole
picture. Such is the nature of our sources that the early years of Osman
remain wrapped in obscurity. When the picture becomes clearer after 1302 it
is possible to follow his relations with Byzantium, but not with his neighbors
or Germiyan or the Mongols. Much remains unknown or unknowable, but I
hope that this view of the Homeland may make its contribution toward see-
ing the rise of the Ottomans in a real context. Yet, as my teacher Sterling
Dow was fond of saying, “there are more questions than answers in this
business.”
Notes
1. Imber 1993, 75.
2. Quoted and translated from APZ by Lindner 2007, 21.
3. See p. 122.
4. See p. 196.
5. See p. 113.
6. For the different types of nomadism, see Khazanov 1994, 19–25. The subject
deserves more study than can be attempted here.
7. Pachymeres III.22 (1.293).
9. Alexander Pope, “An Essay on Criticism” part I, lines 9–10.
10. Lowry 2003, 000 and in considerable detail Kafadar 1999, 29–59; both provide
full references and discuss several other theories that are not treated here.
11. Except by Kafadar 1999, 44f. The work is Togan 1970.
12. On this, see the clear exposition of Fletcher 1979/80.
13. See Lefort 1993, 106–9.
14. See the remarks of Kafadar 1999, 136, and also Fletcher 1979/80, 239, who
explains the importance of tanistry, “the principle of succession that the most
talented male member of the royal clan should inherit the throne.”
15. Imber 1993, 68–71 considers Alaeddin to be entirely fictional.
16. See above p. 152.
17. See Tezcan 2013, who shows that the Tradition eliminated the Mongols, but
that sufficient traces of a different version in the work of APZ indicate a belief
that Turks and Mongols were the original settlers of Asia Minor, before the
Seljuks.
18. Mme Beldiceanu recognizes their importance in her brief sketch of Anatolia in
the thirteenth century: 2003, 355–62, especially 359.
19. See Nicol 1993, 107–40 for a clear account of Byzantium’s troubles in these years.
Bibliography
Sources
Acropolites, George 2007: The History, tr. Ruth Macrides. Oxford.
Aflaki 2002: The Feats of the Knowers of God (Menaqeb al-àrefin), tr. John O’Kane.
Leiden.
Ahmedi 2004: History of the Kings of the Ottoman Lineage and their Holy Raids
against the Infidels, ed. Kemal Silay. Cambridge MA.
Aksarayi 2000: Kerimuddin Mahmud-i Aksarayi Musameretu’l-ahbar, tr. Mürsel
Öztürk. Ankara.
Anonymous 1925: Giese, Friedrich: Die altosmanischen anonymen Chroniken,
Abhandlungen für die Kunde des Morgenlandes 17.1. Leipzig.
Aşıkpaşazade 1929: Die altosmanische Chronik des Asikpasazade, ed. F. Giese (Leipzig
1929; critical text edition); Aşıkpaşaoğlu tarihi, ed. A. N. Atsız (Ankara 1985 and
other printings; modern Turkish); Richard Kreutel, trans., Vom Hirtenzelt zur
hohen Pforte (Graz 1959 (annotated German translation); Aşıkpaşazade Tarihi, ed.
Necdet Öztürk. (Istanbul 2013; facsimile and transcription).
Atsız, C. N., ed. 1961: Osmanlı Tarihine Ait Takvimler. Istanbul.
Barkan, Ömer Lutfi and Enver Meriçli, eds. 1988: Hudavendigar Livası Tahrir
Defterleri I. Ankara.
Battuta, Ibn 1962: The Travels of Ibn Battuta, tr. H. A. R. Gibb, vol. II. Cambridge.
Bibi, Ibn 1959: Die Seltschukengeschichte des Ibn Bibi, tr. Herbert Duda. Copenhagen.
Brocardus 1906: Directorium ad passagium faciendum in Recueil des historiens des
croisades, Documents armeniens 2.367–517. Paris.
Cantacuzene 1828–1832: Ioannis Cantacuzeni historiarum libri IV, ed. Ludwig
Schopen. Bonn.
Chalcocondyles 1923: Laonici Chalcocondylae Historiarum Demonstrationes, ed.
E. Darko. Budapest.
ChronMin 1979: Die Byzantinischen Kleinchroniken, ed. Peter Schreiner. Vienna.
Cinnamus 1836: Ioannis Cinnami epitome rerum ab Ioanne et Alexio Comneno gesta-
rum, ed. August Meineke. Bonn.
Comnena, Anna 1945: Alexiade, ed. Bernard Leib. Paris.
Destan 1954: Le destan d’Umur Pacha, ed. and tr. Irene Melikoff-Sayar. Paris.
Dölger, Regesten 1924–1977: Regesten der Kaiserurkunden des oströmischen Reiches,
ed. Franz Dölger. Munich.
Gregoras, Nicephorus 1828–1830: Byzantina historia, ed. Ludwig Schopen. Bonn.
—1973–2000: Nikephoros Gregoras Rhomäische Geschichte, tr. Jan-Louis van Dieten.
Stuttgart.
Al-Harawi 2004: A Lonely Wayfarer’s Guide to Pilgrimage, ed. Josef Meri. Princeton.
Hopf, Charles 1873: Chroniques greco-romaines. Berlin.
246 Bibliography
MAMA V 1937: Monumenta Asiae Minoris Antiqua V: Monuments from Dorylaeum

and Nacolea, ed. W. M. Cox and A. Cameron. Manchester.
Metochites, Theodore 2007: Hoi dyo basilikoi logoi ed. Ioannis Polemis. Athens.
Miklosich, Franz and Joseph Müller 1860: Acta et diplomata Graeca medii
aevi. Vienna.
Muntaner, Ramon 1926: L’expedicio dels catalans a Orient, ed. Nicolau d’Olwer.
Barcelona.
Neşri 1949: Kitab-I Cihan-Numa Neşri Tarihi, ed. F. R. Unat and M. A. Köymen.
Ankara.
Odo of Deuil 1948: De profectione Ludovici VII in orientem, ed. and tr. V. Berry.
New York.
Pachymeres 1984–1999: Georges Pachymères Relations historiques, ed. Albert
Failler. Paris.
Pachymeres 1984–2000: Georges Pachymeres Relations historiques, ed. Albert
Failler. Paris.
Patriarch Athanasius 1975: The Correspondence of Athanasius I Patriarch of
Constantinople, ed. A.-M. Talbot. Washington DC.
Pegolotti, Francesco Balducci 1936: La pratica della mercatura, ed. Allan Evans.
Cambridge MA.
Rashid al-Din 1998: Jamiʻuʾt-tawarikh = Compendium of chronicles, tr. Wheeler
Thackston. Cambridge MA.
RCEA 1931: Répertoire chronologique d’épigraphie arabe. Cairo.
—1977: Les regestes des actes du patriarchat de Constantinople, ed. J. Darrouzès. Paris.
—2013: Risale-i Felekiyye, ed. Oktay Güvemli, Cengiz Toraman. Istanbul.
Al-Umari 1838: Mesalek alabsar fi memalek alamsar, tr. M. Quatremere in Notices et
extraits des manuscrits de la bibliotheque du roi 13, 151–381.
—1929: Al-Umari’s Bericht über Anatolien in seinem Werke Masalik al-absar fi
mamalik al-amsar, ed. Franz Taeschner. Leipzig.
Uzluk, Feridun, ed. 1952. Histoire des Seldjoukides d’Asie Mineure. Ankara.
Secondary works
Ahrweiler, Helene 1965: “l’histoire et la géographie de la région de Smyne entre les
deux occupations turques (1081–1317),” Travaux et mémoires 1: 1–204.
—1966: Byzance et la mer. Paris.
—1983: “La region de Philadelphie au XIVe siècle,” Comptes redus de l’Académkie des
Inacriptions et Belles Lettres 127–1: 175–97 Album, Stephen 2011: Checklist of
Islamic Coins. Santa Rosa.
Alexandrescu-Dersca, Maria 1977: La Campagne de Timur en Anatolie. London.
Allom, Thomas 1838: Constantinople and the Scenery of the Seven Churches of Asia
Minor illustrated in a Series of Drawings from Nature. London.
Amitai-Preiss, Reuven 1999: “Mongol Imperial Ideology and the Ilkhanid War
against the Mamluks,” in Amitai-Preiss and Morgan 1999, 57–72.
—and David Morgan, edd. 1999: The Mongol Empire and its Legacy. Leiden.
Angold, Michael 1975: A Byzantine Government in Exile. Oxford.
Bibliography 247
Arel, Ayda 1968: “Menteşe beyliği devrinde Peçin şehri,” Anadolu Sanatı
Araştırmaları I:69–101.
Arnakis, G. Georgiades 1947: Hoi Protoi Othomanoi. Athense.
Avramea, Anne 1981: “Manuel Ducas Comnène Gavras de Troade a propos de CIG
IV, no. 8763,” in H. Ahrweiler, ed., Geographica byzantina. Paris: 37–41.
Ayalon, Ami 2012: “Malik,” in Encyclopedia of Islam. 2nd Edition.
Aygil, Yakup and Yakup Özkan 2012: Bithynia Tumluğu Içinde Akmeşe (Armaş).
Istanbul.
Ayverdi, Ekrem Hakki 1974: Osmanli Miʿmârîsinde Fâtih Devri. Istanbul.
Ayverdi, Ekrem Hakki 1966: Osmanlı Mi`marisinin Ilk Devri. Istanbul.
—1972: Çelebi ve II. Sultan Murad Devri. Istanbul.
Babinger, Franz 1922: “Der Islam in Kleinasien,” Zeitschrift der Deutschen
Morgenländischen Gesellschaft 76:126–52.
Baker, Julian 2019: “A reassessment of Wood’s 1871 Artemision hoard of fourteenth
century coins,” Sabine Ladstatter and Paul Magdalino eds., Ephesos from Late
Antiquity until the Late Middle Ages. Vienna: 73–102.
—and Berndt Kluge 2017: “Der Gigliati-schatzfund aus den Ausgrabungen der
Berliner Museen in Milet (1903) verborgen um 1370/74,” Numismatische
Zeitschrift 122/123:367–84.
Başar, Fahmattin 1995: “Ertuğrul Gazi,” Islam Ansiklopedisi.
Beldiceanu-Steinherr, Irene 1965: “La conquete d’Andrinople par les Turcs,” Travaux
et Mémoires 1965:434–61.
—1967, Recherches sur les actes des regnes des sultans Osman, Orkhan et Murad
I. Munich.
—1984: “Notes pour l’histoire d'Alasehir (Philadelphie) au XIVe siècle,” Philadelphie
1984: 17–36.
—2000a: “Pachymère et les sources orientales,” Turcica 32: 425–34.
—2000b: “La conquete de la Bithynie maritime, étape decisive dans la fondation de
l'etat ottoman,” in K. Belke et al., Byzanz als Raum. Vienna: 21–35.
—2002: “Analyse de la titulature d'Orhan sur deux inscriptions de Brousse,” Turcica
34:223–40.
—2003: “L'installation des Ottomans en Bithynie,” in Geyer-Lefort 351–74.
—2015: Etudes ottomano-byzantines. Istanbul.
Belke, Klaus 2007: “Die vier Briefe des Patriarchen Athanasios I. an den Metropoliten
von Apameia und die Eroberung Bithyniens durch die Osmanen am Anfang des
14. Jahrhunderts,” in K. Belke et al., Byzantina Mediterranea, Festschrift fur
Johannes Koder. Vienna: 29–42.
Belon, Pierre 1555: Les observations de Pierre Belon du Mans de plusieurs singulariés.
Antwerp.
Bendall, Simon and Cecile Morrrisson 1979: “Un trésor de ducats d'imitation au
nom d'Andrea Dandolo (1343–1354),” RN ser 6, 21:176–93.
Bertrandon 1892: Le voyage d’Outremer de Bertrandon de la Broquière, ed. Ch. Shefer.
Paris.
Blair, Sheila 1982: “The coins of the later Ilkhanids,” American Numismatic Society
Museum Notes 27:211–30.
Bondoux, Rene 2003: “Les villes,” in Geyer-Lefort 2003: 377–409.
248 Bibliography
Cahen, Claude 1951: “Notes pour l’histoire des Turcomans d’Asie Mineure au XIIIe
siècle,” Journal Asiatique 239:335–54.
—1968: Pre-Ottoman Turkey. London.
—1971: “Questions d’histoire de la province de Kastamonou au XIIIe siècle,”
Selçuklu Araştırmaları Dergisi 3:145–58.
—1974: “Ibn Saìd sur l’Asie-Mineure Seljuqide,” in Turcobyzantina et Oriens
Christianus, Article XI. London.
Çelik, Gülbahar Baran 2011: “Aydos kalesi 2010 yılı kazı çalişmaları,” in U. Demir
and M. Mazak eds., Aydos Kalesi ve Istanbul’un Fethi. Sultanbeyli Belediyesi.
Couroupou, Matoula 1981: “Le siège de Philqdelphie par Umur Pacha,” in
H. Ahrweiler, ed., Geographica byzantina. Paris: 67–77.
Cuinet, V. 1894: La Turquie d’Asie. Paris.
Damali, Atom 2010: Osmanlı Sikkeleri Tarihi/History of Ottoman Coins I. Istanbul.
Danişmend, Ismail Hami 1947: Izahlı Osmanlı tarihi kronolojisi. Istanbul.
Danışman, Zuhuri, ed. 1970: Evliya Celebi Seyahatnamesi. Istanbul.
Dernschwam: Hans Dernschwam, Voyage en Asie Mineure (1555): extracts translated
by Jean-Pierre Grelois in Geyer-Lefort 200: 113–38.
von Diest and Anton 1895: Neue Forschungen im nordwestlichen Kleinasien
(Petermanns Mitteilungen Ergänzungsband 25). Gotha.
von Diest, Walter 1898: Von Tilsitt nach Angora (Petermanns Mitteilungen
Erg-Heft 125).
Diler, Omer 2006: Ilhanids: Coinage of the Persian Mongols. Istanbul.
Doğru, Halime 2001: “Osmanlı kroniklerinde Karacahisar kalesi ve Karacaşehir’in
yer degiştirmesi,” in V. ortacağ ve Türk dönemi kazı ve araştırmaları sempozyumu.
Ankara: 221–33.
—2005: XV. ve XVI. Yüzyılda Eskişehir ve Sultanönü Sancağı. Istanbul.
Dörner, Friedrich Karl 1941: Inschriften und Denkmäler aus Bithynien. Berlin.
van Egmont, Johannes Aegidius 1759: Travels Through Part of Europe, Asia
Minor. London.
Ehlert, Rolf 2014: Umlaufgeld im Osmanischen Reich. Heidelberg.
EI2 1960–2009: The Encyclopaedia of Islam. New Edition. Leiden.
Elisséeff, Nikita 1954: “La titulature de Nur ad-din d'après ses inscriptions,” Bulletin
d’études orientales 14:155–96.
Emecen, Feridun 1995: “Gazaya dair—XIV. Yüzyıl Kaynakları Arasında Bir Gezinti,”
Prof. Dr. Hakki Dursun Yıldız Armağanı, Istanbul: 191–7.
Ender, Celil 1994: Ladik (Denizli) Sikkeleri/Coinage of Ladik (Denizi). Istanbul.
—2000: Karesi, Saruhan, Aydın ve Menteşe Beylikleri Paraları. Istanbul.
—2003: Candaroğulları Beyliği Paraları Kataloğu. Istanbul.
—2005: Germiyanoğulları Beyliği Paraları Kataloğu. Istanbul.
Erdmann, K. 1961: Das anatolische Karavansaray des 13. Jh. Berlin.
Eskişehir İl Yıllığı 1967: Ankara.
Failler, Albert 1990: “Chronologie et composition dans l'histoire de Pachymeres III,”
Revue des études byzantines 48:5–87.
—1994: “Les emirs turcs à la conquete de l’Anatolie au début du 14e siècle,” Revue
des études byzantines 52:69–112.
Bibliography 249
—1996: “Ephèse fut-elle prise en 1304 par les Turcs de Sasan?,” Revue des études byz-
antines 54:245–8.
Fellows, Charles 1852: Travels and Researches in Asia Minor More Particularly in
Lycia. London.
Flemming, Barbara 1964: Landschaftsgeschichte von Pamphylien, Pisidien un Lykien
im Spatmittelalter. Wiesbaden.
Fletcher, Joseph 1979/80: “Turco-Mongolian monarchic tradition in the Ottoman
Empire,” Harvard Ukrainian Studies 3/4: 236–51.
Fontanier, V. 1829: Voyages en orient entrepris par ordre du gouvernement
français. Paris.
Foss, Clive 1976: Byzantine and Turkish Sardis. Cambridge MA.
Foss, Clive 1979: “Late Byzantine fortifications in Lydia,” JOB 28:297–320.
—1982: “The defences of Asia Minor against the Turks,” Greek Orthodox Theological
Review 27:145–205.
—1990: “Byzantine Malagina and the Lower Sangarius,” Anatolian Studies 40:161–83.
—1996a: Survey of Medieval Castles of Anatolia II: Nicomedia. London.
—1996b: “Dorylaion: Bulwark of the Byzantine Frontier,” GOTR 41:39–56.
—1998: “Byzantine responses to Turkish attack: some sites of Asia Minor,” Aetos.
Studies in Honour of Cyril Mango, ed. Ihor Sevcenko and Irmgard Hutter.
Stuttgart.
—2019: “The puzzling coins of the earliest Ottomans and their neighbors,” Revue
numismatique 179:181–201.
—and David Winfield 1986: Byzantine Fortifications, An Introduction. Pretoria.
Fraser, James Baillie 1838: A Winter Journey (Tatar) from Constantinople to
Tehran. London.
French, David 1981: Roman Roads and Milestones of Asia Minor. Oxford.
Gazimihal, Mahmut 1958: “Istanbul Muhasalarında Mihaloğulları ve Fatih Devrine
Ait Bir Vakıf Defterine Göre Harmankaya Malikanesi,” Vaıiflar Dergisi 4:125–36.
Geyer-Lefort 2003: Bernard Geyer and Jacques Lefort eds., La Bithynie au Moyen
Age. Paris.
Gibbons, Herbert Adams 1916: The Foundation of the Ottoman Empire. Oxford.
Giese, Friedrich 1924: “Das Problem der Entstehung des Osmanischen Reiches,”
Zeitschrift. für Semitisik 2:246–71.
Giros, Christophe 2003: “Les fortifications medievales” in Geyer-Lefort 2003,
|209–24.
Goodwin, Godfrey 1971: A History of Ottoman Architecture. London.
von der Goltz, Colmar, Freiherr 1896: Anatolische Ausflüge. Berlin.
Grabar, Andre 1978: Sculptures byzantines du moyan age II. Paris.
Grierson, Philip and Lucia Travaini 1998: Medieval European Coinage 14: Italy (III).
Cambridge.
Grousset, Rene 1970: The Empire of the Steppes. New Brunswick, NJ.
Haci Kalfa (Katip Çelebi) 1852: Cihannuma in Vivien de Saint Martin, Description
historique et géographique de l’Asie Mineure II. Paris: 641–742.
Hamilton, William 1842: Researches in Asia Minor, Pontus and Armenia. London.
Hasluck, F. W. 1906: “Bithynica,” ABSA 13:285–308.
250 Bibliography
—1910: Cyzicus. Cambridge.

Haspels, Emilie 1971: The Highlands of Phrygia. Princeton.
Heyd, W. 1885: Histoire du commerce du Levant au moyen age. Leipzig.
Huart, Cl. 1897: Konia la ville des derviches tourneurs. Paris.
Humann, E. and G. Puchstein 1890: Reisen in Kleinasien und Nordsyrien. Berlin.
Hunger, Herbert 1976: Die hochsprachllche profane Literatur der Byzantiner. Munich.
IA 1988–2013: Turkiye diyanet vakfı Islam Ansiklopedisi. Istanbul.
Imber, Colin 1993: “The legend of Osman Gazi,” in Zachariadou 1993a.67–75.
—1994: “Canon and apocrypha in early Ottoman history,” in C. Heywood and
C. Imber, Studies in Ottoman History in Honour of Professor V. L.Ménage. Istanbul,
117–37.
Inalcık, Halil 1962: “The rise of Ottoman historiography,” in B. Lewis and P. M. Holt,
Historians of the Middle East. London: 152–67.
—1982: “the question of the emergence of the Ottoman State,” International Journal
of Turkish Studies 2:71–9.
—1985: “The Rise of the Turcoman maritime principalities in Anatolia, Byzantium,
and Crusades,” Byzantinische Forschungen 9:179–217.
—1993: “Osman Ghazi's siege of Nicaea and the battle of Bapheus,” in Zachariadou
1993a: 77–99.
—2003: “The struggle between Osman Gazi and the Byzantines for Nicaea,” in
I. Akbaygil, H. Inalcik and O. Aslanapa, edd., Iznik Throughout History, 59–85.
Istanbul.
Izmirlier, Yılmaz 1999: Hamidoğulları Beyliği Paraları. Istanbul.
—2005: Aydınoğulları Beyliği Paraları. Istanbul.
—2009: The Coins of Anatolian Seljuks. Istanbul.
Janin, Raymond 1964: Constantinople Byzantine. Paris.
—1975: Les eglises et les monastères des grands centres byzantins. Paris.
Jones, Christopher 2014: “Louis Robert in Central Mysia,” Chiron 44:23–54.
Kafadar, Cemal 1995: Between Two Worlds, The Construction of the Ottoman State.
Berkeley.
Kaplanoğlu, Raif 2000: Osmanlı Devleti’nin Kuruluşu. Istanbul.
Keppel, George 1831: Narrative of a Journey across the Balcan. London.
Khazanov, Anatoly 1994: Nomads and the Outside World. Madison.
Kinneir, John 1818: Journey through Asia Minor, Armenia and Koordistan. London.
Kiprovska, Mariya 2013: “Byzantine renegade and holy warrior: reassessing the
character of Köse Mihal, a Hero of the Byzantino-Ottoman Borderland,” in
S. Kuru and B. Tezcan, Defterology. Festschrift in Honor of Heath Lowry = Journal
of Turkish Studies 40:245–69.
Kızıltan, A. 1958: Anadolu Beyliklerinde Cami ve Mescitler. Istanbul.
Kolbas, Judith, Timothy May, and Vlastimil Novak 2011: Anatolian Early 14th
Century Coin Hoard. Prague.
Köprülü, M. Fuat 1935: Les origines de l’empire ottoman. Paris.
Korobeinikov, Dmitri 2004: 2004 : “The Revolt in Kastamonu c.1291–1283,”
Byzantinishe Forschungen 28:87–117.
—2014: Byzantium and the Turks in the Thirteenth Century. Oxford.
Körte, Alfred 1895: “Kleinasiatische Studien,” Athenische Mitteilungen 20:1–19.
Bibliography 251
—1897: “Review of G. Radet,” En Phrygie, Gottingische gelehrte Anzeigen

159:386–416.
Kuran, A. 1968: The Mosque in Early Ottoman Architecture. Chicago.
Kyriakides, Savvas 2014: “The revolt of the general Kassianos in Mesothynia (1306),”
Byzantion Nea Hellas 33:165–80.
Laiou, Angeliki 1970: “Marino Sanudo Torsello, Byzantium and the Turks: the back-
ground to the anti-Turkish league of 1332–1334,” Speculum 45:374–92.
Langer, William and Robert Blake 1932: “The rise of the Ottoman Turks and its
historical background,” American Historical Review 37:468–505.
Leake, W. M. 1824: Journal of a Tour in Asia Minor. London.
Lebas Philippe 1845: “Voyage en Asie-Mineure,” Revue de philologie I:27–46.
Lefort, Jacques 1993: “Tableau de la Bithynie au XIIIe siècle,” in Zachariadou
1993: 101–17.
—2003: “Les miniatures de Matrakçı,” in Geyer-Lefort 2003, 99–104.
Lemerle, Paul 1957: L’emirat d’Aydin, Byzance et l’occident. Paris.
—1984: “Philadelphie et l'emirat d'Aydin,” in Philadelphie: 55–67.
Lewis, Bernard 1988: The Poltical Language of Islam. Chicago.
Lindner, Rudi 1983: Nomads and Ottomans in Medieval Anatolia. Bloomington.
—2007: Explorations in Ottoman Prehistory. Ann Arbor.
Lowry, Heath 2003: The Nature of the Early Ottoman State. Albany.
Lucas, Paul 1712: Voyage du Sieur Paul Lucas dans la Grèce, l’Asie Mineure, la
Macédoine et l’Afrique. Paris.
MacFarlane, Charles 1850: Turkey and its Destiny. London.
Mango, Cyfil 1966: “The Monastery of St. Abercius at Kurşunlu (Elegmi) in
Bithynia,” Dumbarton Oaks Papers 22:169–76.
Mansel, Arif Mufid 1936: Yalova und Umgebung. Istanbul.
Mantran, Robert 1954: “Les inscriptions arabes de Brousse,” Bulletin des études
orientales 14:87–114.
Mazarakes, Andreas 2012: To Doukato tes Venetias kai oi Apomimeseis tou sten
Anatole 14–16 AI. Athens.
Melville, Charles 2009: “Anatolia under the Mongols,” in Kate Fleet, ed., The
Cambridge History of Turkey I: 51–101.
Menage, Victor 1962: “The Beginnings of Ottoman Historiography” in B. Lewis and
P. M. Holt, Historians of the Middle East (pp. 168–79). London.
—1963: “The Menaqib of Yakhshi Faqih,” Bulletin of the School of Oriental and
African Studies 26:50–4.
—1964: Nesri’s History of the Ottomans. London.
Mordtmann, A. D. 1925: Anatolien, Skizzen und Reisebriefe aus Kleinasien
(1850–1859), ed. Franz Babinger. Hannover.
Muller-Wiener, W. 1962: “Die Stadtmauer von Izmir, Sığacık und Çandarlı,”
Istanbuler Mitteilungen 12:59–114.
Munro, J. A. R. and H. M. Anthony 1897: “Explorations in Mysia,” The Geographical
Journal 9:150–69.
Naumann, Claudia 1985: “Die mittelalterliche Festung von Aizanoi-Çavdarhisar,”
Istanbuler Mitteilungen 35: 275–94.
252 Bibliography
Necipoğlu, Gülrü 2005: The Age of Sinan: Architectural Culture in the Ottoman
Empire. London.
Neville, Leonora 2018: Guide to Byzantine Historical Writing. Cambridge.
Nicol, Donald 1993: The last centuries of Byzantium, 1261–1453. Cambridge.
Olivier, G. A. 1807: Voyage dans l’Empire Othoman, l’Egypte et la Perse. Paris.
Otter, Jean 1748: Voyage en Turquie et en Perse. Paris.
Otto-Dorn, Katharina 1941: Das islamische Iznik. Berlin.
Parman, Ebru 2001: “Eskişehir-Karacahisar’da bir ortaçağ kalesi,” in V. ortaçağ ve
Türk donemi kazı ve araştırmaları sempozyumu. Ankara: 451–62.
Perk, Haluk, and Hüsnü Öztürk 2008: Eretna Sikkeleri. Istanbul.
Philadelphie 1984: Philadelphie et autres etudes. Paris.
Pitcher, D. E. 1972: A Historical Geography of the Ottoman Empire. Leiden.
Pococke, Richard 1743: A Description of the East. London.
Pralong, Annie 1984: “Les remparts de Philadelphie” in Philadelphie 101–26.
Purchas, S. 1625: Purchas his Pilgrims. London.
Radet, G. 1895: “Rapport sur une mission scientifique en Asie Mineure,” Nouvelles
archives des missions scientifiques et litteraires 6:425–596.
Ramsay, William Mitchell 1890: The Historical Geography of Asia Minor. London.
Refik, A. 1924: “Fatih zamanında Sultan Oyugi,” Turk Tarih Encümeni Mecmuası 14
no. 3(80):129–11.
Reis, Leonhard 2002: “Zur Datierung der lateinischen Prägungen der anatolischen
beyliks im 14. Jahrhundert,” Mitteilungen des Österreichischen Numismatischen
Gesellschaft 42.1:5–13.
Remler, Philip 1980: “Ottoman, isfediyarid, and eretnid coinage: a currency com-
munity in fourteenth century Anatolia,” American Numismatic Society Museum
Notes 25:167–88.
—1985: “New light on economic history from Ilkhanid accounting manuals,” Studia
Iranica 14:157–77.
Riefstahl, Rudolf 1931: Turkish Architecture in Southwestern Anatolia. Cambridge.
Robert, Louis 1962: Villes d’Asie Mineure. Paris.
—1977: “La persistence de la toponymie antique en Anatolie,” in La toponymie
antique. Strasbourg.
—1980: A travers l’Asie Mineure. Paris.
Runciman, Steven 1953: A History of the Crusades I: The First Crusade. Cambridge.
Şahin, Sencer 1981: Katalog der antiken Inschriften des Museums von Iznik (Nikaia)
II.1. Bonn.
Saint-Martin, Vivien de 1846: Histoire des découvertes géographiques. Paris.
Sarman, Özdemir 2001: Bursa Yenişehir 1301–2001. Bursa.
Schreiner, Peter 1969: “Zur Geschichte Philadelpheias im 14. Jahrhundert,”
Orientalia christiana periodica 35:375–431.
Smith, John Masson 1999: “Mongol Nomadism and Middle Eastern geography:
Qīshlāqs and Tümens,” in Amitai-Preiss and Morgan 1999: 39–56.
Spuler, Bertold 1955: Die Mongolen in Iran. Berlin.
Srećković, Slobodan 1999: Akches. Belgrade.
Sümer, Faruk 1969: “Anadolu’da Moğollar,” Selcuklu Araştırmaları Dergisi 1:1–147.
Bibliography 253
Taeschner, Franz 1924–1926: Das anatolische Wegenetz nach osmanischen Quellen.

Leipzig.
—1928: ‘Anatolische Forschungen’, Zeitschrift der Deutschen Morgenländischen
Gesellschaft 82:83–118.
Temir, Ahmet 1953: “Anadoluda Ilhanlı valilerinden Samagar Noyan,” Fuad Köprülü
Armağan. Istanbul: 495–500.
—ed., 1959: Caca oğlu: Kırşehir emiri Caca oğlu Nur el-din’in 1272 Tarihli Arapça-
Moğolca Vakfiyesi. Ankara.
Texier, Charles 1862: Asie Mineure: description geographique, historique et arche-
ologique. Paris.
Tezcan, Baki 2013: “The memory of the Mongols in early Ottoman Historiography,”
in H. Erdem Çıpa and Emine Fetvacı, eds.: Writing History at the Ottoman Court.
Bloomington.
TIB 7 1990: Klaus Belke and Norbert Mersich, Tabula imperii byzantini 7: Phrygien
und Pisidien. Vienna.
Toğan, Zeki Velidi 1970: Umumi Türk Tarihine Giriş. Istanbul.
Tomaschek, W. 1891: Zur historischen Topographie von Kleinasien im Mittelalter =
Sitzungsberichte . . . Wien 124.
Treiyer, Alberto 2017: “The chronology of events in the history of Pachymeres
related to the battle of Bapheus and the beginning of the Ottoman Empire,”
International Journal of Humanities and Social Science 7:23–48.
Tournefort, F. 1717: Relation d’un voyage au Levant fait par ordre du Roy. Paris.
Turan, Osman 1958: Turkiye Selcukluları hakkında resmi vesikalar. Ankara.
—1971: Selçuklular zamanında Türkiye. Istanbul.
Uzunçarşılı, Ismail Hakki 1969: Anadolu Beylikleri. Ankara.
Varlık, Mustafa Cetin 1974: Germiyan-oğullari Tarihi. Ankara.
Vismara, G. 1968: “Le relazioni dell'impero con gli emirati selgiuchidi nel corso del
secolo decimoquarto,” Byzantinische Forschungen 3:210–21.
Vryonis, Speros 1971: The Decline of Medieval Hellenism in Asia Minor and
the Process of Islamization from the Eleventh through the Fifteenth Century.
Berkeley.
Walpole, R. 1820: Travels in Various Countries of the East. London.
Wendel, Carl 1940: “Planudea,” BZ 40:406–45.
Whitby, Michael 1985: “Justinian's Bridge over the Sangarius,” Journal of Hellenic
Studies 105:129–48.
Wiegand, Theodor 1913: Der Latmos (Milet III.1). Berlin.
—1941: Didyma I, 2. Berlin.
Wittek, Paul 1932: “Zur Gescichte Angoras im Mitttelalter,” in Theodor Menzel, ed.,
Festschrift Georg Jacob. Leipzig: 329–54.
—1934: Das Fürstentum Mentesche. Istanbul.
—1938: The Rise of the Ottoman Empire. London.
Wulzinger, Karl et al. 1935: Das islamische Milet. Berlin.
Yerasimos, Stephane 1991: Les voyageurs dans l’Empire Ottomane (XIVe–XVIe
siècles). Ankara.
Yıldırım, Fahri 2006: Sakarya Kaleleri. Adapazar.
254 Bibliography
Yücel, Yaşar 1980: XIII–XV. Yüzyıllar Küzey-batı Anadolu Tarihi Çoban-oulları

Candar-oğulları Beylikleri. Ankara.
Zachariadou, Elizabeth 1983: Trade and Crusade. Venetian Crete and the Emirates of
Menteshe and Aydin (1300–1415). Venice.
—1993a, ed.: The Ottoman Emirate (1300–1389). Rethymnon.
—1993b: “The emirate of Karasi and that of the Ottomans: two rival states,” in
Zachariadou 1993a: 225–36.
—1997: “Gazi Celebi of Sinope,” in Zachariadou, Studies in Pre-Ottoman Turkey and
the Ottomans. Aldershot, Hants. 2007, study IX.
Zhukov, Konstantin 1993: “Ottoman, Karasid, and sarukhanid coinages and the
problem of currency community in Turkish Western Anatolia (‘40s–‘80s of the
Fourteenth Century,” in Zachariadou 1993a: 236–42.
Index
Ab Suyu 69–70 historiography 179–80

Abu Saìd 144, 183, 184, 216, 217, 227 Hızır of Aydın 196, 198, 199, 200, 223
Adranos 62–63, 72, 85 inscriptions 200
Ahmedi 9, 14, 43, 168 Mehmet of Aydın 124–25, 144, 145, 180,
Akça Koca 69, 70, 73, 74 196, 200
Akhisar 136, 166 relations with Germiyan 115, 121, 178
Akyazı 69, 70, 73, 85, 171 in the sources 192, 196–98
Ala ad-Din III 142, 167, 169, 234 territories 163, 165, 166
Ala ad-Din Kayqubad I 10, 157 during the 1330’s 223–24
Alaeddin 158 trade and wealth 198–99
Alaides 111 tribal alliances 109
Alais 111–13, 163–65 see also Umur of Aydın
Alexius I Comnenus 100 Aydos 74–75, 85
Alexius Philanthropenus 107–8 Ayverdi, E. Hakki 3, 6, 26
Alınca 26–27
Alişir of Germiyan 111, 114–15, 127, 163, Balban the Genoese 39, 168, 175, 191, 192,
173–74, 180; see also Germiyan 194, 198, 203, 206, 208, 209, 212,
Alp Arslan 100 214–16, 225
al-Umari 1, 146, 184, 191 Balıkesir 80–83, 148, 150, 205–6, 207, 208,
Ameramanai 111 225; see also Karesi
Amourios 111, 112, 163, 164–65, 176 Bapheus, battle of 99, 110, 137–38, 166
Amourios, Ales 111–13, 176 Baybars 159–60
Andronicus Palaeologus 105–6, 108, 109, Bay Hoca 22, 23, 72
113, 120, 122–24, 177, 182, 207 Beldiceanu-Steinherr, Irene 237
Angelokome 136 Belokome 135–36
Apollonia 81, 82 Bergama 81, 83
Armenocastron 16 Bertrandon de la Broquiere 20
Arnakis, G. G. 236 Beyazit I 23
Arsenius 102 Bilecik
Aşıkpaşazade 2, 5, 9, 46–47, 65, 135–37 Belokome and 135–36
Atares 164–65 castle 24
Atinai see Aydın Osman’s conquest of 57–58, 84
Atmanes 111 site of 23–24
Aya Nikola 22, 23, 136 tekfur’s friendship with Osman 22, 24, 45,
Aydın 57, 87, 164, 166
as Atinai 111 tekfur’s marriage 57–58
attack on Philadelphia 144–45, 177, 178 Bithynia
Birgi 196 under the Byzantines 106, 108, 110–11,
capture of Smyrna 124–25, 224 120–23, 125–26
Cayster valley 199 geographical factors 4
coinage 144–45, 180, 199–200 the Kayi Turks in 10
emirate 185, 218 under the Ottomans 73–74, 126–27
in European sources 198–99 travelers chronicles 4–5
256 index
Bitnos 63 sources on 99

Blake, Robert 236 Tralles 105–6
Bolu Turkish incursions into 108, 109–11
Konur Alp’s conquest of 73, 80, 85
under Ottoman control 85–86 Çadırlı 65–66
site of 52, 54, 85–86 Çakır Pınar 57, 58, 59, 83
Bozüyük 13, 14, 17 Candaroğulları 163, 164–65, 192, 213, 216;
Bursa see also Kastamonu
under the Byzantines 119, 121 caravansarays
coins minted in 81, 143–44, 186 of Eskişehir 40, 41–42
mosque inscription 147 of Mihal Beg 56
Orhan’s conquest of 72–73, 122, 167 of Vezirhan 13
siege 64, 71, 84–85 Catalan Grand Company 113–19, 121,
site of 20, 23 144, 177
in the sources 211 Chalcocondyles 14
Tamerlane’s attack on 14 Choban 39, 174, 183–84, 217, 241
travelers’ descriptions of 17 Çoban Kale 76, 79
Byzantium Çobanoğulları 111, 163, 175
administrative system 87 coins
the Aegean region 102, 109, 111–12 of Ayasuluk 145
alliance with Umur of Aydın 125–26 of Aydın 144–45, 180, 199–200
the Anatolian frontiers 102, 108, 117–21 of Germiyan 146, 209
under Andronicus 105–6 as historical sources 141
Bithynia under 106, 108, 110–11, Ilkhanid 144–46, 241
122–23, 125–26 of Karacahisar 141
Bursa 119, 121 of Karesi 81, 86, 139, 179, 207
Catalan Grand Company 113–19, 121, of Kastamonu 176
144, 177 of Menteşe 144, 182, 195
Cayster valley 105, 107, 115, 145 of the Mongols 145–46, 183, 227–28
Chliara 114 of Orhan in Bursa 81, 143–44, 186
churches 3 of Osman’s reign 141–43
confrontations with the Turks 100–2 of Saruhan 203
diplomacy under Michael of Söğüt 142–43
Palaeologus 104–5, 158–59 Constantinople 101
Duo Bounoi 107 Cotyaeum 100
Iznik 151 Cyzicus 114, 122
Laodicea 100–1
late period monuments 3 Daz Ali 77, 78
Maeander valley 102–3, 105, 107–8 Dernschwam, Hans 16–17, 31, 58–59, 61
Magnesia 115–17 Diest, Major von 33, 49, 50, 55–56
Melangeia 136, 152 Dimboz, campaign of 22, 62–64, 72, 138, 166
Nicaea 100, 101, 103–4, 106, 118, 120 Domaniç Dağ 17–19
Nicomedia 106 Domaniç mountains 11, 17–18, 86–87, 166
Philadelphia 177 Dorylaeum 31, 36–37, 42, 44, 100
relations with the Mongols 103–4, 109, Duo Bounoi 107
119–20, 182 Düzpazar 85
relations with the Seljuks 100–1, 104
Sangarius valley 105, 112–13 Edremit 81
under Saruhan 204–5 Emir Yaman 111
Söğüt 14–15 Enveri 9, 191
index 257
Eretna, Alaeddin 227 as newly established 88, 168, 172–73

Ermene 72–73 Orhan’s titles in inscriptions 147
Ermeni Beli 11, 16–17, 21, 86–87 Philadelphia under 174, 177–79, 180
Ermeni Beli, battle of 22, 23 in the sources 208–10
Ertuğrul during the 1330’s 225–26
conquest of Karacahisar 28 Geyer, Bernard 6
figure of 10 Geyve 68, 69
in Söğüt 11–12, 86 Ghazan 111–13, 142, 146, 160, 161, 174, 182
summer and winter pastures 11, 16–22, Ghiyath ad-Din Kaykhusraw II 158
86–87, 167 Gibbons, Herbert Adams 236
tomb 11, 14, 47 Göl Pazar 55–56, 65, 85, 170
Eskişehir Goltz, General von der 31, 52, 71
Alaeddin mosque 38 Göynük 52, 53, 55, 79, 85, 170, 171,
as a center of trade and 177, 213
agriculture 40, 41–43 Gregoras, Nicephorus 1, 99, 108, 109, 120,
Dorylaeum 31, 36–37, 42, 44, 100 126, 128, 145, 164–65
Jebrail ibn Jaja’s endowment deeds Günyarık 14, 15
38, 40–43, 88, 142
Osman’s conquest of 28, 29, 39, 43–44 Haci Bektaş 167
under the Seljuks/Mongols 38–39, 43, 167 Haci Ilbeg 80, 81
site of 6, 23, 36, 37 Haji Kalfa (Kâtip Çelebi) 13, 31, 34–35, 40
as Sultan Önü 37–38 Hamza Bey 23
as Sultan Öyüğü 37–38 Harmankaya 45, 46, 47–49
as Sultanyuki 38–44, 100 Hasan Bozorg 227
Tamerlane’s attack on 14 Hızır of Aydın 125, 126, 196, 198, 199,
through the reigns of Osman and 200, 223
Orhan 43–44 Hulagu, sultan 104, 113, 120, 158
in the tradition 168 Humann, Karl 17, 25
travelers’ descriptions of 5–6, 10, 40–41
Evliya Çelebi 11, 14, 20, 40, 52, 61, 63, Ibn Battuta 1, 39, 49, 50, 52, 53–54, 55, 85,
67–68, 75 148–49, 150, 171, 175, 186, 191, 193,
196–97, 202, 205, 208, 209, 211–12, 213,
First Crusade forces 15, 37, 100 214–15, 232
Fourth Crusade 101 Ibn Saìd 38, 176–77, 182
Ilkhanids
Gaikhatu 161 association with Kastamonu 176–77
Gazi Mihal 45, 55 coinage 144, 145–46, 241
Gazi Mihal Beg 45–46 Gaikhatu 161
Gemlik 72 Hulagu, sultan 104, 113, 120, 158
Gengis Khan 157, 184 Jebrail ibn Jaja’s endowment deeds
Gerede 177, 192, 213–14, 226 38, 40–43, 88, 142
Germiyan Mamluke threat to 159
Alişir of Germiyan 111, 114–15, 127, 163, Ottoman relations with 88
173, 174, 180 relations with the Ottomans 241
alliances with other chiefs 109, 173–74 relations with the Seljurks 160–61
coinage 146, 209 relations with the Turks 184–85
emirate of 173–74, 185, 208–11 sultanate 104
enmity with Osman 29, 32, 174–75, 241 territories 44
at Karacahisar 29, 32–33, 170 during the 1330’s 227
Kütahya headquarters 44, 87, 173, 208, 209 Inalcık, Halil 6, 31, 137, 237
258 index
İnegöl Kafadar, Cemal 237

Angelokome and 136 Kaikhusraw III 38, 44
tekfur’s enmity with Osman 22, 25, Kale Tepe 76
87, 266 Kandall 27, 34, 44
town 22–23, 84 Kandıra 73
İnegöl, plain of Karacahisar
as a Byzantine base 22–23 castle 31–32, 44
Osman’s crossing of 18–20, 22 collection of taxes in 32, 60
summer pastures 16, 17, 19, 86–87 conquest of 28, 29, 32–33, 87, 141,
travelers’ descriptions of 17, 20, 21 167, 168–70
İnönü Germiyan at 29, 32–33, 170
fortified cave 34–35, 36 location of 30–31
mosque of Yadigar 35 Osman’s base in 48, 85, 169
under Orhan 61, 168 Osman’s proclamation of
site of 13, 34–35 independence 28, 29, 60–61, 166
in the tradition 167–68, 177 passed to Orhan 30, 169
inscriptions striking of coins by Osman 141
of Aydın 200 Tatar attack on 30
in Menteşe 195 in the tradition 168–70
on mosques 146–48 Karacaşehir 31, 33
Orhan’s titles in 146–48 Karamanids 159–60
Izmit Kara Tekin (Karadin) 70–71
conquest of 75–76, 86 Karesi
Osman’s campaign to reach 69–71 Akira 206
in the tradition 138 Balıkesir 80, 81, 82, 83, 148, 150, 205–6,
see also Nicomedia 207, 208, 225
Iznik Bergama 81, 83, 205
under the Byzantines 151 coinage 81, 86, 139, 179, 207
conquest of 85, 86 Edremit 81
conversion to Islam 77 in European sources 207–8
as Osman’s capital 6, 77–79 Orhan’s conquest of 80–82, 86, 138–39
siege of 76–78 relations with the Byzantines 122
Tamerlane’s attack on 14 in the sources 192, 205–6
in the tradition 138 territories 165, 179
see also Nicaea during the 1330’s 225
Izz ed-Din 112, 158, 159, 160 Yahşi Khan 205, 206, 207
Kastamonu
Jebrail ibn Jaja 38, 40–43, 88, 142, 167, 168 during the 1330’s 227
John Comnenus 63, 82, 100 under Ales Amourios 112
John Palaeologus 102–3 emirate of 146, 160, 165, 175–77,
John Tarchaniotes 108 214–16
John VI Cantacuzene mosques 148
alliance with Alişir of Germiyan 127 Suleyman Paşa 39
alliance with Orhan 126–27, 226 Kâtip Çelebi (Haci Kalfa) 13, 31, 34–35, 40
alliance with Umur of Aydın 125–27, 201 Kayi Turks 10
campaign in Bithynia 122–23 Kestel 63
defence of Thrace 126–27 Kirmasti 82
memoirs 1, 99–100 Kite 63
Orhan’s threat to Nicomedia 124 Konur Alp
on Osman’s tribal alliances 164, 234 Akyazı 69, 70, 73, 85, 171
Justinian’s bridge 70, 71, 83, 84 Bolu 73, 80, 85
index 259
Konurapa 85 emirate of 163, 180–82, 185

Mudurnu 73, 80, 85, 170, 171 in European sources 194–95
Konurapa 80 in inscriptions 195
Köprü Hisar 62 island fortress Duo Bounoi 107
Köprülü, Fuat 236 as Mantakhiai 111
Köprülü Mehmet Pasha 13 naval power 110
Köse Mihal Orhan’s titles in inscriptions 147–48
family and historicity of 45–47, 57, 169, 234 Peşin 148, 149, 193
predatory raids across the Sangarius in the sources 192, 193–94
55, 56, 85, 87 territories 164, 165, 166
Koyunhisar 84 during the 1330’s 223
Kulaca 25–26, 166 trade and wealth 195–96
Kuleler 66 Michael IX 109, 111
Kütahya Michael Palaeologus
Germiyan headquarters 44, 87, 173, defence of Paphlagonia 175–76
208, 209 defence of Philadelphia 177
as a major Ottoman site 6, 100 diplomacy of 104–5, 113, 158–59
medrese 121 as emperor 102
route to Bursa 20 on the Turcomans 101
Mihal Beg of Gölpazar 45–46, 56
Lamises 111, 164 Mihaliçi 82, 83
Langer, William 236 Mihalloğulları 46, 47
Laodicea 100–1, 108 Mongols
Lascarids 101 Abu Saìd 144, 183, 184, 216, 217, 227
Leblebicihisar 65 Choban 39, 174, 183–84, 217, 241
Lefke 65 coinage 145–46, 183, 227–28
Lefort, Jacques 6 control of the Anatolian plateau 159, 217
Leo Muzalon 110 defeat of Osman 120, 137, 138, 241
Lindner, Rudi 6, 237 in European sources 217
Lopadion 122 Ghazan 111–12, 113, 142, 146, 160, 161,
Lowry, Heath 235, 237 174, 182–83
Golden Horde 105, 109, 157, 158
Macdonald Kinneir, John 4, 20 Hulagu, sultan 104, 113, 120, 158
MacFarlane, Charles 21, 61, 62, 63 in Philadelphia 122
Malagina 66–67, 152, 166, 172 presence in the Homeland 88
Mal Hatun 10, 28 relations with Osman 172
Mamlukes 158, 159; see also Ilkhanids relations with the Byzantines 103–4, 109,
Mantakhiai see Menteşe 119–120, 182
Manuel Comnenus 20, 31, 37, 44, 100 relations with the Ottomans 241
Manzikert, battle of 100 relations with the Seljurks 101, 157–60,
Maria (daughter of Michael VIII) 104, 167, 182
113, 120 relations with the Turks 111–13
Mehmed the Conqueror 43–44 in the sources 216–17
Mehmet of Aydın 124–25, 144, 145, 180, territories 157–58, 216–17
196, 200 during the 1330’s 227–28
Mekece 66, 151, 172 Timurtash 39, 122, 168, 178, 180, 184,
Melangeia 136, 152 217, 227; see also Ilkhanids
Menteşe Mordtmann, Andreas David 5–6, 21, 23, 61
capture of Tralles 105–6 mosques
coins 144, 182, 195 Alaeddin mosque, Eskişehir 38
collaboration with other chiefs 109 built by Orhan 150
260 index
mosques (cont.) mosques of 150

of Bursa 147 in Nicomedia 123–24
construction of, fourteenth century 148–49 Orhan’s fortress, Bilecik 22, 24
great mosque of Birgi 148, 149, 180 Orhan’s mosque, Bilecik 24–25
inscriptions on 146–48 as padishah 81
mosque of Orhan, Yarhisar 58, 60 rule over Karacahisar 30, 169
mosque of Yadigar, İnönü 35 in the sources 211–12
Orhan’s mosque, Bilecik 24–25 in the suburbs of Constantinople 85
Ulu Cami 147 territorial expansion 226, 235,
Mudurnu 238–40, 241–42
Konur Alp’s conquest of 73, 80, 85, 170 titles of in inscriptions 146–48
under Ottoman control 79, 85 in the tradition 234, 235
raids on 45, 48 in vakf (endowment) documents 150–52
site of 54–55 Osman
travelers’ descriptions of 52–53, 54 battle of Bapheus 99, 110, 137–38, 166
Murat I 1, 39, 44, 82, 144, 168 in Byzantine sources 109
campaign in Izmit 69–71
Neşri 2, 9, 10, 141, 167, 168 campaign of Dimboz 22, 62–64, 72,
Nicaea 138, 166
under the Byzantines 100, 101, 103–4, coins of 141–43
106, 118, 120 conquest of Bilecik 57–58, 84
churches 3 conquest of Eskişehir 28, 29, 39, 43–44
conquest of 1, 84 conquest of Karacahisar 28–29, 48, 60–61,
the First Crusade in 15 85, 166, 169
Lascarid tower 78 conquest of the Sangarius
tower of Andronicus II 106 valley 65–68, 84, 166
the Turks in 100, 110, 120, 122, 123, conquests against the Byzantines 1
137–38, 167 date of the death of 186–87
view down on 77 early sources on 1–2
see also Iznik emergence of the Ottomans under 1–2, 84
Nicomedia enmity of the tekfur of İnegöl 22, 25, 87, 166
under the Byzantines 106 enmity with Germiyan 29, 32, 174–75, 241
fall to the Ottomans 110, 126, 138, 167 friendship with the tekfur of Bilecik 22, 24,
site of 125, 126 45, 57, 164, 166
threat from the Turks 123–24 in İnegöl 18–20, 22
see also Izmit Iznik as a capital of 77–79
Niketiaton 123 marriage to Mal Hatun 10, 28
Mongol defeat of 120, 137, 138, 241
Orhan Pachymeres’s history of 99
alliance with John Cantacuzene 127, 226 predatory raids across the Sangarius 45, 48,
in Bithynia 122–23 50, 55, 56–57, 85
campaign against Izmit 69, 75–77 proclamation of independence
capture of Bursa 72–73, 122, 167 28, 29, 60–61
capture of Nicomedia 126 relations with the Christian
coins of 81, 143–44, 186 population 22, 29, 45, 166
conquest of Karasi 80–82, 86, 138–39 relations with the Mongols 172
conquest of Kara Tekin 69 in Scutari 110, 136
date for the accession of 186–87 territorial expansion 166–72, 186, 233–35,
early sources on 1 237–38, 241–42
İnonu under 61 title of Fakhr al-din 142, 169, 234
index 261
in vakf (endowment) documents 150–52 Philanthropenos 178

victory at Koyunhisar 84 poetry 9, 191
Yenişehir as the capital of 61, 150, Prusa see Bursa
166, 234
Ottoman emirate Qawiya 214
emergence of 1, 234–43
naval power 110, 212 Roger de Flor 114–17, 119
Osman’s proclamation of Rukneddin 158
independence 28, 29, 60–61
in the sources 192, 211–12 Sagoudaous 135
during the 1330’s 226 Salampaxides 111; see also Menteşe
Turcoman emirates 217–18 Samandıra 73, 74
Ottoman history Samsa Çavuş 45, 53, 55–57, 65, 66, 85,
challenges of reconstruction 231 170, 234
chronology of conquests 171 Samsun 214
comparison of sources 135–39 Sangarius valley
early sources on 1–2, 5, 191 Ab Suyu 69–70
European sources 191–92 Akyazı 69, 70, 73, 85, 171
geographical factors 3–4, 87–88 Bozaniç Kaya 50, 51
the Homeland 231–32, 233 under the Byzantines 105, 112–13
monuments of 3 Çadırlı 65–66
Osman’s territorial expansion 166–72, Geyve 68, 69
186, 233–35, 237–38, 241–42 Göl Pazar 55–56, 65, 85, 170
scholarship on 5–6, 231, 235–37 Göynük 52, 53, 55, 79, 85, 170, 171,
significant sites 3–4 177, 213
Turkish sources 2 Justinian’s bridge 70, 71, 83, 84
validity of the early sources 2–4, 83–87 Kara Tekin (Karadin) 70–71
Ottoman prehistory 2, 9 Leblebicihisar 65
Oynaş 33–34 Lefke 65
Malagina 66–68, 152, 166, 172
Pachymeres, George 1, 99, 100, 101–2, 107, Mekece 66, 151, 172
108, 111, 115, 118, 120, 135–38, 144, Osman’s conquest of 65–68, 84, 166
145, 163, 166, 172, 180, 233 Osman’s predatory raids in 45, 48, 50, 55,
Pagdinai 111 56–57, 85
Paphlagonia 103, 109, 110, 111, 138, region of 49–50, 52–54
163, 175–76 Taraklı 45, 50–52, 55, 85
Pazarcık 16–17 White Castle of the Sangarius 67, 68
Pegolotti, Francesco 192, 202, 207, 209 Sardis 112
Peşin 148, 149, 193 Saruhan
Philadelphia alliance with Umur of Aydın
under the Byzantines 102, 105, 107, 125, 126, 127
114–16, 177 under the Byzantines 204–5
fall to the Ottomans 128 Cantacuzene’s attacks on 127
under the Germiyan 174, 177–79, 180 coinage 203
the Mongols in 122 historiography 179
siege of 121, 177, 178 mosques 149
in the sources 202 Nymphaeum 203, 204
during the 1330’s 224 in the sources 192, 202–3
Umur of Aydın’s attack on 125, 127–28, territories 165, 166
201, 202, 224 during the 1330’s 224–25
262 index
Saru Yatı 26 Taeschner, Franz 6

Sasa 109, 115, 144–45, 164–66, 174, 180 Tamerlane 14
Sasan 120, 145, 163 Tarakçı Yenicesi 79, 85, 170, 171
Scutari 110, 136 Taraklı 45, 50–52, 55, 85
Second Crusade 17, 37 Tatars 30, 33, 169
Seljurks Texier, Charles 20, 21, 61, 71
Ala ad-Din II 28 Thrace 9, 125, 126–27, 136, 157, 191, 200,
Ala-ed- Din Kayqubad I 10 201, 204, 207, 223, 224, 225
Eskişehir under 38–39, 43, 167 Timurtash 39, 122, 168, 178, 180, 184,
Mamluke threat to 159–160 217, 227
Mongol defeat of 157–58 Togan, Zeki Velidi 236
Osman’s proclamation of Tralles 105–6
independence 28 Tripolis 114, 115
relations with the
Byzantines 100–1, 104 Ulubat 63–64, 80–82
relations with the Ilkhanids 160–61 Umur of Aydın
relations with the Mongols 101, 157–60, alliance with John Cantacuzene 125–26,
167, 182 127, 201
territories 157 alliance with Saruhan 125, 126, 127
Sfondylai 111, 164 attack on Philadelphia 125, 127–28, 201,
Sinope 101, 151, 157, 163, 175, 215, 216 202, 224
Söğüt the Dusturname 9, 191
in Byzantine sources 14–15 naval power 127, 201
coins 142–43 during the 1330’s 223
in the epic poem of Ahmedi 14 as Umar of Izmir 201–2
as the first homeland 10–11, 86, 167 al-Uryan, Haydar 85, 171, 191–93, 197, 202,
geography of 11, 12–13, 22 206, 208, 212, 213, 215
Sagoudaous and 135
tomb of Ertuğrul 11, 14 vakf (endowment) documents 150–52
as a transport hub 10, 13–16
as winter pastures 22, 86 Wittek, Paul 236
Sülemis 182
Süleyman Paşa 39, 52 Yahşı Fakih 9, 46–47, 126–27, 144, 225
Süleymanşah 10, 35 Yalova 75–76, 86
Sultan Önü 28, 37, 42, 44, 214, 217; Yarhisar
see also İnönü; Eskişehir; conquest of 84, 166
Karacahisar importance of 58–59
Sultanönü 28, 34, 37, 168, 177, 213–14; mosque of Orhan 58, 60
see also Eskişehir Yenişehir
al-Sultanoyugi, Mahmud ibn Mehmed 39 as a major Ottoman site 6, 58, 85
Sultan Öyüğü 28, 37–38, 40, 167, 168 as Osman’s new capital 61, 150, 166, 234
Sultanyuki 38–44, 100 plain of 61–62
Süleyman 75, 79, 85 Tamerlane’s attack on 14

The Beginnings of The Ottoman Empire Clive Foss 2022

Uploaded by

Copyright:

Available Formats

The Beginnings of The Ottoman Empire Clive Foss 2022

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Beginnings of The Ottoman Empire Clive Foss 2022

Uploaded by

Copyright:

Available Formats

OUP CORRECTED PROOF – FINAL, 15/12/21, SPi

OXFORD STUDIES IN BYZANTIUM

Church Architecture of Late Antique Northern Mesopotamia

The Beginnings of the

1.1 Ertuğrul’s tomb in Söğüt 11

xii List of Illustrations

1.29 Justinian’s bridge over the Sangarius 70

List of Illustrations xiii

7.8 Nymphaeum: the Lascarid palace 204

1. The Homeland

List of Maps xvii

3. The Marmara region

2 The Beginnings of the Ottoman Empire

4 The Beginnings of the Ottoman Empire

In treating this material, I have used a chronological arrangement, with

6 The Beginnings of the Ottoman Empire

deserve a more thorough investigation by an historian; until now, however,

8 The Beginnings of the Ottoman Empire

10 The Beginnings of the Ottoman Empire

Ottoman Prehistory (APZ cap. 2)

The Homeland of the Ottomans 11

The First Settlement (APZ cap. 2)

Fig. 1.1 Ertuğrul’s tomb in Söğüt

12 The Beginnings of the Ottoman Empire

Fig. 1.2 Söğüt, seen from the southeast

Fig. 1.3 Söğüt, from the west

The Homeland of the Ottomans 13

14 The Beginnings of the Ottoman Empire

Söğüt thus has a curiously dual character: it is a remote mountain town

The Homeland of the Ottomans 15

Fig. 1.4 Byzantine remains at Günyarık

century.21 This thus appears to be a structure rebuilt by the Lascarids and

16 The Beginnings of the Ottoman Empire

The Summer Pastures

The Homeland of the Ottomans 17

18 The Beginnings of the Ottoman Empire

The Homeland of the Ottomans 19

Fig. 1.7 The great forest near Domaniç

20 The Beginnings of the Ottoman Empire

The Homeland of the Ottomans 21

22 The Beginnings of the Ottoman Empire

First Battles and Conquests (APZ caps. 3, 5)

The Homeland of the Ottomans 23

24 The Beginnings of the Ottoman Empire

Fig. 1.8 The site of Bilecik

The Homeland of the Ottomans 25

Fig. 1.9 Bilecik: Orhan’s mosque

“1286” First Victory (APZ cap. 5)

26 The Beginnings of the Ottoman Empire

“1286” (APZ cap. 5)

The Homeland of the Ottomans 27

28 The Beginnings of the Ottoman Empire

Karacahisar, Sultanönü, and Conflict

The conquest of Karacahisar in 1288 looms large in Ottoman tradition.

The Homeland of the Ottomans 29

30 The Beginnings of the Ottoman Empire

Fig. 1.10 Karacahisar barely visible on top of the hill [10]

The Homeland of the Ottomans 31

32 The Beginnings of the Ottoman Empire

control of Ankara, under the real or nominal suzerainty of Ala ad-Din